From sebestyen_g at accentcomm.com Sat May 1 00:38:47 1999 From: sebestyen_g at accentcomm.com (Gabor Sebestyen) Date: Thu Nov 3 14:42:01 2005 Subject: OS4.2i/Ethernet: URGENT! Multicast-aware drivers ... Message-ID: <372AAF87.C99EE2E7@accentcomm.com> Dear Great Old OPENSTEP NetExperts, NeXTSTEP/OPENSTEP systems can serve virtual hosts, if ps_if kernel driver is installed ... and having an ethernet card whose driver supports multicast. We recently changed our PC server to a bigger (Celeron 366) one and the old ethernet card which was good before now under heavy packet traffic locks up the machine. We decided to get another one but we failed because those probed ones didn't support multicasting that is there weren't virtual hosts. We tried Intel EtherExpress Pro and DECnet card but noone was able to serve virtual hosts. Only the old AMDPCnet one but it kills the machine when "packet attack" occurs. So my questions are: Am I right? It is true what I wrote? Who uses or used OPENSTEP server as virtual host provider and which ethernet driver is/was used? Please help us because Monday is so close ... Thank you in advance! Best regards, Gabor Sebestyen From sebestyen_g at accentcomm.com Sat May 1 11:26:19 1999 From: sebestyen_g at accentcomm.com (Gabor Sebestyen) Date: Thu Nov 3 14:42:01 2005 Subject: Develope Driver for OS4.2 Intel Message-ID: <372B474B.500A4140@accentcomm.com> Hi, I know that officially no way to develope driver on OS4.2. Anyway in reality how can I do it? Best regards, G?bor -------------------------------------------------------------- A C C E N T C O M M U N I C A T I O N S - We Make Your Web Groove - Tel: (+36 1) 315-0701 http://www.accentcomm.com (+36 1) 343-4440 e-mail: info@accentcomm.com GSM: (+36 30) 951-4334 Fax: (+36 30) 980-0344 -------------------------------------------------------------- From sebestyen_g at accentcomm.com Sun May 2 07:44:57 1999 From: sebestyen_g at accentcomm.com (Gabor Sebestyen) Date: Thu Nov 3 14:42:01 2005 Subject: Develope Driver for OS4.2 Intel References: <372B474B.500A4140@accentcomm.com> Message-ID: <372C64E9.2E0B6A6B@accentcomm.com> Yeah! I managed to recompile a driver in OS4.2 ... :-) I do not know how I did but ... it's working! Great! Gabor From sebestyen_g at accentcomm.com Sun May 2 08:04:02 1999 From: sebestyen_g at accentcomm.com (Gabor Sebestyen) Date: Thu Nov 3 14:42:01 2005 Subject: OS4.2i/Ethernet: part II. References: <372AAF87.C99EE2E7@accentcomm.com> Message-ID: <372C6962.FCD1D364@accentcomm.com> Well! There was a suprise! I tried the same card which hadn't been working in our server home and it was working ... At ISP ps_if displayed the following message: "May 1 14:05:11 next3 mach: Invalid ps control add-multicast" At home on the same OS (OS4.2 Intel + Patch2), and with same driver and card there wasn't problem at all!. It was working! Now I really don't undestand what's happening... The next step is to guess what is the difference between the two systems. Any idea, suggestion? Gabor -- -------------------------------------------------------------- A C C E N T C O M M U N I C A T I O N S - We Make Your Web Groove - Tel: (+36 1) 315-0701 http://www.accentcomm.com (+36 1) 343-4440 e-mail: info@accentcomm.com GSM: (+36 30) 951-4334 Fax: (+36 30) 980-0344 -------------------------------------------------------------- From owolf at advancenet.net Sun May 2 09:12:42 1999 From: owolf at advancenet.net (Scott Johnson) Date: Thu Nov 3 14:42:01 2005 Subject: ObjC LOC (lines of code) counter?? Message-ID: <199905021514.KAA22650@babba.advancenet.net> Would a better measure of productivity be the number of milestones reached in your project plan? You do have a written software requirements document with concrete milestones layed out on some sort of time chart? Granted a resource dependency chart rather than a raw time chart is a more accurate measure, but most customers don't understand that projects can get delayed because of a dependency upon another component rather than "low productivity" Scott Johnson ========================================================================== ORANGE WOLF SOFTWARE ========================================================================== From eric at alum.mit.edu Sun May 2 13:10:34 1999 From: eric at alum.mit.edu (Eric Hermanson) Date: Thu Nov 3 14:42:01 2005 Subject: Problems with setAutosaveTableColumns: and NSDocument Message-ID: <372CB139.3C7C8727@alum.mit.edu> Hello, I'm using YellowBox on WebObjects 4 / NT and I'm having two problems. The first is that attempting to use -[NSTableView setAutosaveTableColumns:YES] in order to save the UI state of a table view in the defautls database fails with the following exception: -[NSConcreteMutableArray addObject:]: attempt to insert nil #0 0x320411e0 in -[NSException raise] () #1 0x320413b7 in +[NSException raise:format:] () #2 0x320114e3 in _NSArrayRaiseInsertNilException () #3 0x32014bc0 in -[NSMutableArray addObject:] () #4 0x342132b6 in -[NSTableView _readPersistentTableColumns] () #5 0x342131af in NSTableView.o:__i_NSTableView__setAutosaveTableColumns_ () #6 0x403e02 in -[TRDocument windowControllerDidLoadNib:] (self=0xaf6300, ... I call [tableView setAutosaveName:@"MyTVName"]; [tableView setAutosaveTableColumns:YES]; in the - (void)windowControllerDidLoadNib: method of my NSDocument subclass. The exception gets raised as the nib is loaded. Note that this table view is hooked up to an EODisplayGroup via several column associations. I'm thinking that the identifier isn't set correctly or something? The other problem is that I have the document's undo manager hooked to the main EOEditingContext's undo manager. I assumed this would automatically cause the document to mark itself as edited when the user changes a value, and it would mark itself as un-edited if the user reverts. While the edited==YES status is being set correctly, if I click revert: on the editing context, the document still remains edited. Is this a bug in AppKit or EOF? Thank You, Eric From malte at oops.se Mon May 3 01:33:29 1999 From: malte at oops.se (Malte Tancred) Date: Thu Nov 3 14:42:01 2005 Subject: Subprojects in a Legacy project Message-ID: Hi, how can I make subprojects in a legacy project available from PB? When porting standard unix software to MOSXS I like to work with PB. The problem I've encountered is making sourcecode in subdirectories of the unix source available as subprojects in PB? How do you do to achieve this? Do you turn all subdirs to PB subprojects? Do you perhaps create a symlink named *.subproj to the subdirs? Cheerio, Malte -- Malte Tancred, OOPS ab malte@oops.se http://www.oops.se From marco at sente.ch Mon May 3 05:34:18 1999 From: marco at sente.ch (Marco Scheurer) Date: Thu Nov 3 14:42:01 2005 Subject: ObjC LOC (lines of code) counter?? In-Reply-To: <199904301458.IAA30530@soren.lanl.gov> References: <199904301458.IAA30530@soren.lanl.gov> Message-ID: <199905031234.OAA08651@xlthlx.sente.ch> > >Oops. Hit the wrong key. I was just going to point out that it >depends on how heavily uses the Foundation classes. If you just >roll your own code it probably only 1.1 time C. One language comparison (Capers Jones, see http://www.spr.com/) gives the following numbers for programming language levels and ranges of source code statement per function point: Level Min Mode Max C 2.5 60 128 170 C++ 6.0 40 55 140 Objective-C 12.0 17 27 38 Smalltalk 15.0 12 21 30 > Apparently > studies have shown that OO progamming (outside of GUI stuff) > hasn't led to productivity enhancements. I argue that that > is because the right components haven't been built yet for other > programming domains. Other factors: - learning curve, - multiple, diverse environments, some good, some bad, - cost and value of reuse not properly evaluated. A common trend on many oo projects I've seen (including OPENSTEP and Smalltalk projects) was to "generalize" to the point of rewriting foundation classes and development tools, which of course does not help productivity. Marco Scheurer Sen:te From cmh at greendragon.com Mon May 3 06:14:02 1999 From: cmh at greendragon.com (Chris Hanson) Date: Thu Nov 3 14:42:01 2005 Subject: class-dump for Mac OS X Server? Message-ID: I've poked around a bit, and can't find it. Does it even exist? From bbum at codefab.com Mon May 3 07:20:33 1999 From: bbum at codefab.com (Bill Bumgarner) Date: Thu Nov 3 14:42:01 2005 Subject: class-dump for Mac OS X Server? References: Message-ID: <372DB0B1.CC68C305@mail.codefab.com> [I can send you a binary, if you need] Version 2.1 built fine on RDR2 and Mac OS X-- you should be able to find it at next-ftp.peak.org b.bum Chris Hanson wrote: > I've poked around a bit, and can't find it. Does it even exist? From eric at alum.mit.edu Mon May 3 08:55:03 1999 From: eric at alum.mit.edu (Eric Hermanson) Date: Thu Nov 3 14:42:01 2005 Subject: EOUserDefaultsAdaptor? Message-ID: <372DC6D7.FA0760D0@alum.mit.edu> Has anyone thought about writing an EOAdaptor that is based on the NSUserDefaults database? I think this might be desirable for the following reasons: - It would allow the use of an EOModel to model things like user preferences, etc. It could even be used as a poor-man's database system, much nicer than using the Flat File Adaptor. And since NSUserDefaults can be overridden by supplying defaults in higher level domains (like the argument domain), one could even override the data to be fetched on the command-line at runtime. - Since the defautls database is self-creating, there is no need to worry about managing a database -- the adaptor would just call get/set methods on the shared NSUserDefaults instance. - The defaults database stores representations of property list objects like arrays and dictionaries, so it is, in a way, like an object-oriented database. If this adaptor were built, the connection dictionary might contain the defaults system username, the entity name could be the defaults domain name, and the attributes could represent the key/value pairs in the defaults database. The only disadvantages I can think of are in the areas of performance (does anyone know how fast reading/writing from NSUserDefaults is)? - Eric Hermanson From jjfeiler at relief.com Mon May 3 12:01:49 1999 From: jjfeiler at relief.com (John Jay Feiler) Date: Thu Nov 3 14:42:01 2005 Subject: EOUserDefaultsAdaptor? Message-ID: <199905022303.QAA00129@haque.relief.com> > > Has anyone thought about writing an EOAdaptor that is > based on the NSUserDefaults database? I think this might > be desirable for the following reasons: > > - It would allow the use of an EOModel to model things > like user preferences, etc. It could even be used as a > poor-man's database system, much nicer than using the > Flat File Adaptor. And since NSUserDefaults can be > overridden by supplying defaults in higher level domains > (like the argument domain), one could even override the > data to be fetched on the command-line at runtime. > > - Since the defautls database is self-creating, there is > no need to worry about managing a database -- the adaptor > would just call get/set methods on the shared > NSUserDefaults instance. > > - The defaults database stores representations of property > list objects like arrays and dictionaries, so it is, in > a way, like an object-oriented database. > > If this adaptor were built, the connection dictionary > might contain the defaults system username, the entity > name could be the defaults domain name, and the attributes > could represent the key/value pairs in the defaults > database. > > The only disadvantages I can think of are in the areas > of performance (does anyone know how fast reading/writing > from NSUserDefaults is)? > > - Eric Hermanson > Intruiging idea. Someone out there built an LDAP EOAdaptor, and I think that the ability to use the EOInterface layer for the UI of a preferences panel could be pretty efficient. I prototyped a PropertyListEOAdaptor a while back, which worked with ASCII plists, streams, and NSPPLs, and the performance was relatively good. Doing the same thing with NSUserDefaults seems quite reasonable. J -- John Feiler jjfeiler@relief.com From pierce at twinforces.com Mon May 3 12:21:52 1999 From: pierce at twinforces.com (Pierce T. Wetter III) Date: Thu Nov 3 14:42:01 2005 Subject: EOUserDefaultsAdaptor? In-Reply-To: <199905022303.QAA00129@haque.relief.com> References: <199905022303.QAA00129@haque.relief.com> Message-ID: At 12:10 PM -0700 5/3/99, John Jay Feiler wrote: > > > > Has anyone thought about writing an EOAdaptor that is > > based on the NSUserDefaults database? Or even better, how about an EOAdaptor for NetInfo? Pierce ---------------------------------------------------------------- Pierce T. Wetter III, Director, Twin Forces, Inc. e-mail: pierce@twinforces.com Phone:520-779-4227 U.S. Mail: 1300 South Milton Rd, Suite 206, Flagstaff, AZ 86001 From mtarbell at tarbell.jpl.nasa.gov Mon May 3 13:28:02 1999 From: mtarbell at tarbell.jpl.nasa.gov (Mark Tarbell) Date: Thu Nov 3 14:42:01 2005 Subject: Terminal i/o routines Message-ID: <199905032028.NAA07982@tarbell.jpl.nasa.gov> Anyone know where these terminal i/o routines reside in MacOSXS or OS4.2? /bin/ld: Undefined symbols: _tcgetattr _tcsetattr _cfgetispeed _cfgetospeed _cfsetispeed _cfsetospeed *** Exit 1 They're in the man pages, but I can't find the library / framework I need to link against (this is a 'tool'). I realize it's possible they might not exist on MacOSXS, but they should at least be on OS4.2... Thanks Mark -- Mark Tarbell (Mark.A.Tarbell@jpl.nasa.gov) Applications Development Section Information Systems Development and Operations Division Engineering and Science Directorate Jet Propulsion Laboratory 4800 Oak Grove Drive MS 502-500 Pasadena, CA 91109 Opinions expressed herein do not reflect those of JPL. From sanguish at digifix.com Mon May 3 14:52:08 1999 From: sanguish at digifix.com (Scott Anguish) Date: Thu Nov 3 14:42:01 2005 Subject: HTMLEditor - Series cancelled Message-ID: <199905032149.RAA12864@digifix.com> As some of you may have read on Stepwise, this series is now defunct. For those who aren't aware... HTMLEdit was an application and article series started on Stepwise on July 8th of last year. The initial idea was to demonstrate how trivial it is to build a multi-window multi-document application using Apple's newly introduced NSDocument family of classes. This is exactly the kind of thing that, in my opinion, Stepwise should be doing. And I (along with a small band of editors) have been trying to do this for the past few years since Apple aquired NeXT. When Apple released the beta of WebObjects 4.0 in the early fall of 1998, they included a demo application called HTMLEdit. This conflicted with the name of my article series and application. I informed Apple of the conflict as soon as I became aware of the it. Unfortunately, comments to Apple at the time yielded no result. When Apple released WebObjects 4.0, they included the still un-renamed HTMLEdit demo application. This left me in an akward situation. Users were asking me for support, and for the source code for Apple's HTMLEdit, and congratulating me about the inclusion with WebObjects, and that wasn't my application. At no time has Apple EVER asked me to rename my series, nor has anyone at Apple commented on this issue in any sort of an official capacity. As well, I had an unfinished article series that now needed renaming and substantial editing, and seven code archives that needed to be updated. More difficult was the fact that this series had been aimed at the novice, and as such contained a large number of graphics, all of which had HTMLEdit placed throughout them. A proper correction of this would have required redoing all the graphics as well. Lets face it, suing Apple over something like this isn't going to solve anything... what I really wanted was a "sorry about the trouble we've caused you, and thanks for writing the articles to begin with". However, that isn't going to happen. All this hanging over my head (as far as continuing the series), a lingering frustration about the lack of response by Apple, and the hassle of renaming all of the existing articles, changing and testing all the code archives, and re-doing all the graphics with the new name proved to be more than I was willing to deal with. All this during a period where Apple was totally silent about Mac OS X Server. Today (May 3, 99) I got email from yet another person asking me about HTMLEdit (Apple's, not mine), and I took the time to do a partial conversion of the articles. I've renamed the article files and graphics on the server and I've done a search and replace in the actual text of the HTMLEditor articles. I've also modified the source archives and repackaged them, I don't think I've missed anything, but its possible. I did not make new graphics for all the screenshots. That means that there are places where HTMLEdit is referenced in the screenshot, but HTMLEditor is listed in the text. Hopefully inexperienced developers will be able to work around this inconsistency. Putting out a series of articles like this is very tough. It takes substantially longer to track and document the development process than it does to actually write the code. Taking the screenshots, ensuring that they are clearly labelled and accurate to the article itself is a major pain in the butt. I had considered re-starting the series in the recent months, but given the state of the first seven articles (with the incorrect graphics), that its no longer very 'novice' oriented and a general malais about the situation with Apple, I've decided not to continue the series. I continue to use an enhanced HTMLEditor in house to maintain Stepwise, however I'm unsure that it'll ever be in a position where its documented well enough to be used by others. This naming issue was the death of the series... I guess its been on life-support since November, but I'm just now officially pulling the plug. The existing articles and source, in their semi-inconsistent state, will remain on Stepwise for the forseeable future. It bothers me greatly that the graphics don't match up with the new HTMLEditor name, but for the moment at least, I'll live with that. They've moved to a new location on Stepwise, more consistent with the current name. http://www.stepwise.com/Articles/Technical/HTMLEditor To those who enjoyed the article series, I'm glad. I've heard that some developers have used these materials in training courses for clients and others using them as training material for new hires. The series does still have value, even in its current inconsistent state. That's nice to know, and its good that someone has gained some benefit from them.. It is possible that I'll start another article series in the future, but I've got some wounds to lick on this particular issue. Scott Anguish www.stepwise.com -- Scott Anguish Middlebury College - Center for Education Technology Stepwise Server - Mac OS X Server Information From paulrs at lgs-systems.com Mon May 3 12:52:01 1999 From: paulrs at lgs-systems.com (Paul R. Summermatter) Date: Thu Nov 3 14:42:01 2005 Subject: EOUserDefaultsAdaptor? In-Reply-To: <199905022303.QAA00129@haque.relief.com> References: <199905022303.QAA00129@haque.relief.com> Message-ID: <9905031952.AA01354@slab> Guys, What would be much nicer is if Apple would open up the API to allow user defaults to be saved to a real database. For some reason, Apple didn't seem to have admitted the possibility of more than one writable domain. It would have been nice to be able to create a writable domain and put it in the list ahead of the application domain. In this way, we could easily restore a domain from a database when a user logs into the database, and synchronize it when the user logs out, leaving the application domain untouched. As it stands right now, we've had to create our own domain and our own defaults manager which means that any of the built in default functionality (like saving window positions) aren't written to our domain. As for using the defaults as a database of its own, I can't imagine what kind of Registry bloat nightmare this would create on NT, especially with roaming profiles. Regards, Paul You wrote: [ > [ > Has anyone thought about writing an EOAdaptor that is [ > based on the NSUserDefaults database? I think this might [ > be desirable for the following reasons: [ > [ > - It would allow the use of an EOModel to model things [ > like user preferences, etc. It could even be used as a [ > poor-man's database system, much nicer than using the [ > Flat File Adaptor. And since NSUserDefaults can be [ > overridden by supplying defaults in higher level domains [ > (like the argument domain), one could even override the [ > data to be fetched on the command-line at runtime. [ > [ > - Since the defautls database is self-creating, there is [ > no need to worry about managing a database -- the adaptor [ > would just call get/set methods on the shared [ > NSUserDefaults instance. [ > [ > - The defaults database stores representations of property [ > list objects like arrays and dictionaries, so it is, in [ > a way, like an object-oriented database. [ > [ > If this adaptor were built, the connection dictionary [ > might contain the defaults system username, the entity [ > name could be the defaults domain name, and the attributes [ > could represent the key/value pairs in the defaults [ > database. [ > [ > The only disadvantages I can think of are in the areas [ > of performance (does anyone know how fast reading/writing [ > from NSUserDefaults is)? [ > [ > - Eric Hermanson [ > [ [ Intruiging idea. Someone out there built an LDAP EOAdaptor, and I [ think that the ability to use the EOInterface layer for the UI of a [ preferences panel could be pretty efficient. [ [ I prototyped a PropertyListEOAdaptor a while back, which worked with [ ASCII plists, streams, and NSPPLs, and the performance was [ relatively good. Doing the same thing with NSUserDefaults seems [ quite reasonable. [ [ J [ [ -- [ John Feiler [ jjfeiler@relief.com [ --- Paul Summermatter LGS Systems, Inc. Medical Computing Division 15 TJ Gamester Ave Portsmouth, NH 03801-5871 (603) 433-9822 voice (603) 433-9818 fax (888) 898-6321 pager 8986321@skytel.com paging email paulrs@lgs-systems.com (NeXT or MIME Mail Welcome) http://www.lgs-systems.com From paul at infront.com Tue May 4 05:03:17 1999 From: paul at infront.com (Paul Naro) Date: Thu Nov 3 14:42:01 2005 Subject: Adaptec 2940UW Driver for Mac OS X Server Message-ID: <1286303098-24404863@tlww.com> An old post from April 27th ... for those interested > What are the chances of seeing an Adaptec > 2940UW driver for Mac OS X Server? A response from my inquiry to Adaptec ... Dear Paul: According to our Technical Support department, Adaptec does anticipate to eventually release drivers for Mac OS X Servers. At this time, we do not have any dates that we expect to release these drivers. You can visit Apple's web site and compatibility list to see if they are currently offering any drivers for the Power Domain AHA-2940UW Mac Kit and the Mac OS X Server. If you have any further questions, please do not hesitate to contact us. From cmh at greendragon.com Tue May 4 06:36:38 1999 From: cmh at greendragon.com (Chris Hanson) Date: Thu Nov 3 14:42:01 2005 Subject: class-dump for Mac OS X Server? In-Reply-To: <372DB0B1.CC68C305@mail.codefab.com> Message-ID: Thanks to everyone for the info. I found the sources to class-dump in the NeXT developer section on Peak. I haven't had time to build yet, but it doesn't seem like it'll be a problem. From rasmussn at soren.lanl.gov Tue May 4 01:22:18 1999 From: rasmussn at soren.lanl.gov (Craig E Rasmussen) Date: Thu Nov 3 14:42:01 2005 Subject: ObjC LOC (lines of code) counter?? In-Reply-To: <199905031234.OAA08651@xlthlx.sente.ch> (message from Marco Scheurer on Mon, 3 May 1999 05:49:37 -0700 (PDT)) References: <199905031234.OAA08651@xlthlx.sente.ch> Message-ID: <199905040822.CAA00712@soren.lanl.gov> >One language comparison (Capers Jones, see http://www.spr.com/) gives the >following numbers for programming language levels and ranges of source code >statement per function point: > > Level Min Mode Max >C 2.5 60 128 170 >C++ 6.0 40 55 140 >Objective-C 12.0 17 27 38 >Smalltalk 15.0 12 21 30 This is intriguing. What allows a language to reach a higher level? Is it library support? Does static/runtime binding matter? I noticed that Java is level 6.0 as is C++. I would have expected Java to be much better than C++ (about equivalent with Objective-C). I'm clueless. Craig From dondi at itmedicine.net Tue May 4 09:51:46 1999 From: dondi at itmedicine.net (John David N. Dionisio) Date: Thu Nov 3 14:42:01 2005 Subject: More pure Java + PB questions Message-ID: Hi all, Thanks very much for your initial tips on getting a pure Java build going with Project Builder. I'm happy to report that my initial attempts have been successful since getting the pointers, and now I'm ready for more :) Question 1: Our project includes image files (.gifs and .jpgs) in various directories throughout our Java hierarchy. We have been packaging these files with the .jar files we produce and use assorted getResource() methods to access them. Can ProjectBuilder be made to include such image files in its .zip as well? I know that it can include non-source files in its build, but my main challenge is how to get the files into specific directories. PB currently infers the directory hierarchy in my symbolic linked code using the package declarations in the source, but images won't have that. Is there a way to do this, or have we hit a Java-PB impedance mismatch barrier? Question 2: PB's builds are currently from the ground up, not quite "smart." Each build command recompiles the entire Java source hierarchy (which, on MOSXS and a G3/300, is tolerable, but definitely not topflight efficient particularly if you're just concentrating on one or two classes). Can the Makefile(s) be tweaked to build a bit more intelligently? Or, alternatively, is there a way to individually compile .java files? I've noticed that there is such a button in PB, but that it's disabled, most likely because it expects Objective C code. Can this be worked around somehow? Alright, that's it for the moment. I hope I'm not barking up the wrong tree by using PB for something that it wasn't entirely intended to do (i.e. build pure Java apps). I'd have stuck with CodeWarrior on Mac OS and MRJ 2.1.1, which have, of late, actually become compatible and stable enough for development, but our project's class names are going waaaaaaaay beyond the Mac OS's 32-character filename limit! John David N. Dionisio, PhD "All great ideas look like bad UCLA Telemedicine Division ideas to people who are losers." --- Dilbert, 11/18/97 From pcoskren at bbn.com Tue May 4 10:05:58 1999 From: pcoskren at bbn.com (Patrick Coskren) Date: Thu Nov 3 14:42:01 2005 Subject: More pure Java + PB questions Message-ID: <199905041705.NAA22421@po1.bbn.com> From: "John David N. Dionisio" > Question 1: > > Our project includes image files (.gifs and .jpgs) in various directories > throughout our Java hierarchy. We have been packaging these files with the > ..jar files we produce and use assorted getResource() methods to access > them. > > Can ProjectBuilder be made to include such image files in its .zip as well? > I know that it can include non-source files in its build, but my main > challenge is how to get the files into specific directories. PB currently > infers the directory hierarchy in my symbolic linked code using the package > declarations in the source, but images won't have that. > > Is there a way to do this, or have we hit a Java-PB impedance mismatch > barrier? I don't know the direct answer, but an idea for a workaround. PB spits out a zip file. You could have a shell script that uses jar to manually place each of your images into the archive. You might even be able to invoke this script from the Makefile.postamble so it happens automatically, although I don't know how you'd do it. > Alright, that's it for the moment. I hope I'm not barking up the wrong > tree by using PB for something that it wasn't entirely intended to do (i.e. > build pure Java apps). I'd have stuck with CodeWarrior on Mac OS and MRJ > 2.1.1, which have, of late, actually become compatible and stable enough > for development, but our project's class names are going waaaaaaaay beyond > the Mac OS's 32-character filename limit! That shouldn't matter, unless the filenames of the .java files differ past the 32nd character (class files don't matter). Even then you can work around it by renaming some of the files. The reason is that CodeWarrior completely ignores the filename when it's compiling; it gets all the information that it needs from the "package" line at the top of your source file. On the output side, just compile directly to jar. That way you never spit out .class files, and the class name length doesn't matter. At least, that's how it's _supposed_ to work. :-) Actual experience may differ. -Patrick From maury at OAAI.COM Tue May 4 15:08:23 1999 From: maury at OAAI.COM (Maury Markowitz) Date: Thu Nov 3 14:42:01 2005 Subject: NSUndo and document dirtying Message-ID: <199905042205.SAA20597@OAAI.COM> I don't know if this is my fault or not so I'll see if anyone else has seen something similar. Our main view supports Undo. However after making changes and then undoing a lot of times, the document never becomes clean again. I thought it was "duh" code on my part, because I was updating the change counts in places that didn't make sense. However I removed that code and the same thing happens. Overriding updateChangeCount turned all dirtying off as expected, but that's about it. Anyone? How does the document _get_ updateChangeCount from the "do" anyway, does the undo manager call it directly? Maury From zmonster at heliosgroup.com Tue May 4 15:25:11 1999 From: zmonster at heliosgroup.com (Eric Hermanson) Date: Thu Nov 3 14:42:01 2005 Subject: NSUndo and document dirtying Message-ID: <5948F7268092D211B2020000D11C0FB8081802@chi-serv2.heliosgroup.com> I've had trouble with this too. I don't think the NSUndoManager/NSDocument combo correctly cleans the window's document edited state (in WebObjects 4). I think you have to explicitly do this in your code (in your case, your view delegate). I know for sure this doesn't work when using EOExtensions/EODocument in WebObjects 4. The problem there is that the -[EOEditingContext revert] sends it's objects changed in editing context notification too early and it's 'hasChanges' flag is still yes when the notification is sent, so you have to tell the window to clean it's dirty state using performSelector:afterDelay:... Eric > -----Original Message----- > From: Maury Markowitz [mailto:maury@OAAI.COM] > Sent: Tuesday, May 04, 1999 5:13 PM > To: Multiple recipients of list > Subject: NSUndo and document dirtying > > > I don't know if this is my fault or not so I'll see if anyone else > has seen something similar. Our main view supports Undo. However > after making changes and then undoing a lot of times, the document > never becomes clean again. > > I thought it was "duh" code on my part, because I was updating the > change counts in places that didn't make sense. However I removed > that code and the same thing happens. Overriding updateChangeCount > turned all dirtying off as expected, but that's about it. > > Anyone? How does the document _get_ updateChangeCount from the > "do" anyway, does the undo manager call it directly? > > Maury > From sanguish at digifix.com Tue May 4 15:33:42 1999 From: sanguish at digifix.com (Scott Anguish) Date: Thu Nov 3 14:42:01 2005 Subject: Anyone using QT on OSXS? Message-ID: <199905042233.SAA18327@digifix.com> Is there anyone out there besides me that is having a stab at using the limited NSMovieView stuff on OSXS? I'm having horrible problems with crashes and exceptions and all sorts of general ugliness... Are there any other solutions available? From maury at OAAI.COM Tue May 4 15:59:58 1999 From: maury at OAAI.COM (Maury Markowitz) Date: Thu Nov 3 14:42:01 2005 Subject: NSUndo and document dirtying Message-ID: <199905042256.SAA20681@OAAI.COM> > I've had trouble with this too. I don't think the NSUndoManager/NSDocument > combo correctly cleans the window's document edited state (in WebObjects 4). > I think you have to explicitly do this in your code This is my thinking too, but then I see apps that do work properly and I wonder. The rest of the system works properly though, the undo stack gets to zero and beeps as you'd expect, but it's still dirty. Clearly the undo stack is proper, but it never seems to count correctly. Maury From chad at objectwerks.com Tue May 4 16:36:18 1999 From: chad at objectwerks.com (Chad Leigh, ObjectWerks, Inc.) Date: Thu Nov 3 14:42:01 2005 Subject: singleton objects and dealloc Message-ID: <2428610245.925846578@aleigh.cac.stratus.com> Hi I have a class that is a singletone object class. There is only ever one instance and the factory methods and init methods always return that one instance. That instance is declared static id foo; in my class .m file. I want to know what the class -dealloc should do with regards to that instance (or another method if not dealloc). Should it do anything? I am kind of thinking now that it shouldn't do anything. Those who have retained it will release it... Thanks for any comments on how to do singleton objects Chad Chad Leigh, VP Engineering ObjectWerks, Inc. Enterprise Solutions Provider and Consulting chad@objectwerks.com From bbum at codefab.com Tue May 4 16:42:03 1999 From: bbum at codefab.com (Bill Bumgarner) Date: Thu Nov 3 14:42:01 2005 Subject: singleton objects and dealloc In-Reply-To: <2428610245.925846578@aleigh.cac.stratus.com> Message-ID: I would recommend overriding -release/-autorelease and doing absolutely nothing in 'em. Alternatively, you could test to see if the operation would destroy the object and indicate to the programmer that they have just committed a fowl (which they definitely have-- if you retain the singleton once during singleton creation, the programmer should never have code that sends a -release that would cause it to die!). b.bum On Tue, 4 May 1999, Chad Leigh, ObjectWerks, Inc. wrote: > Hi > > I have a class that is a singletone object class. There is only ever one > instance and the factory methods and init methods always return that one > instance. > > That instance is declared > > static id foo; > > in my class .m file. > > I want to know what the class -dealloc should do with regards to that > instance (or another method if not dealloc). Should it do anything? I am > kind of thinking now that it shouldn't do anything. Those who have retained > it will release it... > > Thanks for any comments on how to do singleton objects > Chad > > Chad Leigh, VP Engineering > ObjectWerks, Inc. > Enterprise Solutions Provider and Consulting > chad@objectwerks.com > > From bungi at omnigroup.com Tue May 4 16:48:14 1999 From: bungi at omnigroup.com (Timothy J. Wood) Date: Thu Nov 3 14:42:01 2005 Subject: singleton objects and dealloc Message-ID: <199905042348.QAA09839@mungi.omnigroup.com> A non-text attachment was scrubbed... Name: not available Type: text/enriched Size: 1551 bytes Desc: not available Url : /mailman/archive/macosx-dev/attachments/19990504/c9f161f6/attachment.bin From pierce at twinforces.com Tue May 4 16:57:27 1999 From: pierce at twinforces.com (Pierce T. Wetter III) Date: Thu Nov 3 14:42:01 2005 Subject: singleton objects and dealloc In-Reply-To: <2428610245.925846578@aleigh.cac.stratus.com> References: <2428610245.925846578@aleigh.cac.stratus.com> Message-ID: At 4:44 PM -0700 5/4/99, Chad Leigh, ObjectWerks, Inc. wrote: >Hi > >I have a class that is a singletone object class. There is only ever one >instance and the factory methods and init methods always return that one >instance. > >That instance is declared > >static id foo; > >in my class .m file. > >I want to know what the class -dealloc should do with regards to that >instance (or another method if not dealloc). Should it do anything? I am >kind of thinking now that it shouldn't do anything. Those who have retained >it will release it... > >Thanks for any comments on how to do singleton objects >Chad It depends on how you plan on using the singleton. Should it ever be deallocated at all? Most of the time when I use a singleton, I want it to be a persistent singleton In which case, your factory should retain it, and dealloc isn't an issue. If its a "non persistent" singleton, then the users should retain it, because otherwise it will get autoreleased on the next autorelease cycle. As for your dealloc should do, it should do whatever it would do normally if it wasn't a singleton. Pierce ---------------------------------------------------------------- Pierce T. Wetter III, Director, Twin Forces, Inc. e-mail: pierce@twinforces.com Phone:520-779-4227 U.S. Mail: 1300 South Milton Rd, Suite 206, Flagstaff, AZ 86001 From leigh at ernst.tomandandy.com Tue May 4 17:28:08 1999 From: leigh at ernst.tomandandy.com (Leigh Smith) Date: Thu Nov 3 14:42:01 2005 Subject: Documentation hierarchy in frameworks Message-ID: <199905050028.UAA03199@ernst.tomandandy.com> I'm unable to figure out how to incorporate a documentation hierarchy of similar size and structure to the FoundationKit in a framework, so that installation of the framework will include documentation. Initially the documentation (for the MusicKit which has been ported to MOXS) is in RTF, but in the future would end up as html. Any suggestions? Thanks in advance Leigh -- Leigh Smith leigh@tomandandy.com (MIME) tomandandy +1-212-334-0421 (W) +1-212-334-0422 (F) 89 Greene St. New York, NY 10012, USA http://www.cs.uwa.edu.au/~leigh Microsoft - What do you want to re-install today? From nberch at db.lv Wed May 5 02:58:15 1999 From: nberch at db.lv (Nils Berzins) Date: Thu Nov 3 14:42:01 2005 Subject: Anyone using QT on OSXS? Message-ID: <199905051000.NAA14269@saulite> >Is there anyone out there besides me that is having a stab at using the >limited NSMovieView stuff on OSXS? Hi, Sorry for the stupid question but what's the version of QuickTime that shipped with MacOS X Server ? Or more appropriate question would probably be, what functionality does it have comparing to QT 3.0 for MacOS ? I haven't yet had the time to check it out myself and most likely will have some time only after WWDC. Server finally arrived to my mailbox yesterday :-) Best Regards, Nils Berzins From lavoie at cst.ca Wed May 5 05:59:26 1999 From: lavoie at cst.ca (Martin-Gilles) Date: Thu Nov 3 14:42:01 2005 Subject: Anyone using QT on OSXS? Message-ID: <199905051259.IAA26223@plexus.cst.ca> >>Is there anyone out there besides me that is having a stab at using the >>limited NSMovieView stuff on OSXS? > >Hi, > >Sorry for the stupid question but what's the version of QuickTime >that shipped with MacOS X Server ? Or more appropriate question >would probably be, what functionality does it have comparing to >QT 3.0 for MacOS ? I haven't yet had the time to check it out >myself and most likely will have some time only after WWDC. >Server finally arrived to my mailbox yesterday :-) It's based on QT 3, but without many CODECs like MPEG, AIFF, WAV and then some. The release note in /System/Documentation/ReadMe/Programs&Applications/QuickTime.html has the details. Martin-Gilles Lavoie | "No! Try not. Do, or do not. Mac OS programmer, CS&T | There is no try." lavoie@cst.ca | -- Yoda on error handling http://blackhole.cst.ca/ | Y2K statement: hahahaha! From ullius at ivt.baum.ethz.ch Wed May 5 06:54:32 1999 From: ullius at ivt.baum.ethz.ch (Markus Ullius) Date: Thu Nov 3 14:42:01 2005 Subject: LNK2001 Errors on WebOjects 4.0 for NT Message-ID: Help! I have to port my Application "OpenTimeTable" to WindowsNT 4. In Mac OS X Server I have no problems. When I build my App on NT I get the error-messages listed below. I followed the conversion instructions for the makefiles. I created a new project which compiled successful. But when I added my programfiles to this project the same problem came back. What can I do???? Thanks Markus Ullius Here is a part of the output: ..... e:/programme/Apple/Developer/Libraries/libNSWinMain.a OTTController.o : error LNK2001: unresolved external symbol __imp__NSHomeDirectory OTTDocument.o : error LNK2001: unresolved external symbol __imp__NSHomeDirectory OTTImportController.o : error LNK2001: unresolved external symbol __imp__NSHomeDirectory OTTDayRunning.o : error LNK2001: unresolved external symbol __imp__NSApp OTTGraphicModel.o : error LNK2001: unresolved external symbol __imp__NSApp OTTAnalyzer.o : error LNK2001: unresolved external symbol __imp__NSApp ..... ..... LRegress.o : error LNK2001: unresolved external symbol .objc_class_name_NSAssertionHandler OpenTimeTable_main.o : error LNK2001: unresolved external symbol __imp__NSApplicationMain drawWraps.o : error LNK2001: unresolved external symbol __imp__DPSGetCurrentContext libNSWinMain.a(NSWinMain.o) : error LNK2001: unresolved external symbol __imp___NSArgv libNSWinMain.a(NSWinMain.o) : error LNK2001: unresolved external symbol __imp___NSArgc libNSWinMain.a(NSWinMain.o) : error LNK2001: unresolved external symbol .objc_class_name_NSApplication E:/OpenTimeTable/OTT/OpenTimeTable/OpenTimeTable.app/OpenTimeTable.exe : fatal error LNK1120: 69 unresolved externals make: *** [E:/OpenTimeTable/OTT/OpenTimeTable/OpenTimeTable.app/OpenTimeTable.exe] Error 1 ------------------------------------------------------------------ Markus Ullius Institute of Transportation, Traffic, Highway & Railway Engineering (IVT) ETH Hoenggerberg 8092 Zurich Switzerland Fax: +41 1 633 10 57 Phone: +41 1 633 31 05 Direct: +41 1 633 24 15 -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: text/enriched Size: 2108 bytes Desc: not available Url : /mailman/archive/macosx-dev/attachments/19990505/3ac3c93e/attachment.bin From chad at objectwerks.com Wed May 5 07:31:22 1999 From: chad at objectwerks.com (Chad Leigh, ObjectWerks, Inc.) Date: Thu Nov 3 14:42:01 2005 Subject: singleton objects and dealloc In-Reply-To: Message-ID: <2482314247.925900282@aleigh.cac.stratus.com> b.bum (CodeFab) and Tim Wood (omnigroup) both gave similar answers to this. I am adding a subclass (of an abstract class) to an existing application where other subclasses exist and are used (a type of inspector). Mine needs to be a singleton object. I guess if I retain it once at creation then no matter what the other people using it do it should work out, as long as they practice good memory discipline. Overriding the release mechanism as outlined below and in tim's post would make this easier and more sure. I don't think I care whether it is persistent or not, meaning always the same one, as long as there is never more than one around. But making it a persistent one would make it easier. Thanks all! Chad --On Tuesday, May 04, 1999, 7:42 PM -0400 Bill Bumgarner wrote: > > I would recommend overriding -release/-autorelease and doing absolutely > nothing in 'em. Alternatively, you could test to see if the operation > would destroy the object and indicate to the programmer that they have > just committed a fowl (which they definitely have-- if you retain the > singleton once during singleton creation, the programmer should never have > code that sends a -release that would cause it to die!). > > b.bum > > On Tue, 4 May 1999, Chad Leigh, ObjectWerks, Inc. wrote: > >> Hi >> >> I have a class that is a singletone object class. There is only ever one >> instance and the factory methods and init methods always return that one >> instance. >> >> That instance is declared >> >> static id foo; >> >> in my class .m file. >> >> I want to know what the class -dealloc should do with regards to that >> instance (or another method if not dealloc). Should it do anything? I >> am kind of thinking now that it shouldn't do anything. Those who have >> retained it will release it... >> >> Thanks for any comments on how to do singleton objects >> Chad >> >> Chad Leigh, VP Engineering >> ObjectWerks, Inc. >> Enterprise Solutions Provider and Consulting >> chad@objectwerks.com >> >> > Chad Leigh, VP Engineering ObjectWerks, Inc. Enterprise Solutions Provider and Consulting chad@objectwerks.com From ullius at ivt.baum.ethz.ch Wed May 5 07:42:17 1999 From: ullius at ivt.baum.ethz.ch (Markus Ullius) Date: Thu Nov 3 14:42:01 2005 Subject: LNK2001 Errors on WebOjects 4.0 for NT SOLVED Message-ID: After receiving some hints I checked out my project once again: The AppKit.framework and Foundation.framework are located in the Frameworks-suitcase. They appear also in the makefile. I included them once again in the makefile.preamble et voil? - it worked! I don't know why I have to do this twice but it works.... Thanks you all for your help. Best Regards Markus ------------------------------------------------------------------ Markus Ullius Institute of Transportation, Traffic, Highway & Railway Engineering (IVT) ETH Hoenggerberg 8092 Zurich Switzerland Fax: +41 1 633 10 57 Phone: +41 1 633 31 05 Direct: +41 1 633 24 15 From marco at sente.ch Wed May 5 08:19:42 1999 From: marco at sente.ch (Marco Scheurer) Date: Thu Nov 3 14:42:01 2005 Subject: ObjC LOC (lines of code) counter?? In-Reply-To: <199905040822.CAA00712@soren.lanl.gov> References: <199905040822.CAA00712@soren.lanl.gov> Message-ID: <199905051519.RAA09837@xlthlx.sente.ch> > >>One language comparison (Capers Jones, see http://www.spr.com/) gives the >>following numbers for programming language levels and ranges of source code >>statement per function point: >> >> Level Min Mode Max >>C 2.5 60 128 170 >>C++ 6.0 40 55 140 >>Objective-C 12.0 17 27 38 >>Smalltalk 15.0 12 21 30 > >This is intriguing. What allows a language to reach a higher level? >Is it library support? Does static/runtime binding matter? In fact library support should be taken into account in other entries. For instance, Symantec C++ is rated higher level than plain C++. Runtime binding is certainly a factor in making Smalltalk a higher level language than C++, it is such a simple and powerful thing to use. >I noticed that Java is level 6.0 as is C++. I would have expected >Java to be much better than C++ (about equivalent with Objective-C). > >I'm clueless. I agree, this is strange. My version of the table had no entry for Java. The one on the web is dated of 1996. I suspect it is either a case of missing data, and (bad?) extrapolation (this would cast a doubt on the whole comparison) or because of the immaturity of the language at the time. Marco Scheurer Sen:te From josh at i33.com Wed May 5 09:09:31 1999 From: josh at i33.com (Josh Flowers) Date: Thu Nov 3 14:42:01 2005 Subject: PB and Java packages. Message-ID: <199905051609.MAA22774@fnord.i33.com> Is it possible to separate classes by package in PB? I tried creating subprojects for each package, but when I compiled, the subprojects appear to get compiled separately, which is problematic when classes in separate subprojects refer to each other. Is there a better way to work with java packages in PB, or are subprojects the accepted way? From josh at i33.com Wed May 5 09:09:21 1999 From: josh at i33.com (Josh Flowers) Date: Thu Nov 3 14:42:01 2005 Subject: No subject Message-ID: <199905051609.MAA22763@fnord.i33.com> Is it possible to separate classes by package in PB? I tried creating subprojects for each package, but when I compiled, the subprojects appear to get compiled separately, which is problematic when classes in separate subprojects refer to each other. Is there a better way to work with java packages in PB, or are subprojects the accepted way? From josh at i33.com Wed May 5 09:25:56 1999 From: josh at i33.com (Josh Flowers) Date: Thu Nov 3 14:42:01 2005 Subject: PB and Java packages. Message-ID: <199905051625.MAA23765@fnord.i33.com> My mistake. I just tried a test case, and it worked. The initial problem must have been mine. > s it possible to separate classes by package in PB? I tried > creating subprojects for each package, but when I compiled, the > subprojects appear to get compiled separately, which is problematic > when classes in separate subprojects refer to each other. Is there a > better way to work with java packages in PB, or are subprojects the > accepted way? From josh at i33.com Wed May 5 09:46:26 1999 From: josh at i33.com (Josh Flowers) Date: Thu Nov 3 14:42:01 2005 Subject: PB and Java packages. Message-ID: <199905051646.MAA25124@fnord.i33.com> I just ran a test, and PB does work properly with subpackages. Apparently I made a different mistake earlier (and my MailViewer skill don't seem much better, sorry about the duplicates). > Is it possible to separate classes by package in PB? I tried > creating subprojects for each package, but when I compiled, the > subprojects appear to get compiled separately, which is problematic > when classes in separate subprojects refer to each other. Is there a > better way to work with java packages in PB, or are subprojects the > accepted way? From pcoskren at bbn.com Wed May 5 09:50:11 1999 From: pcoskren at bbn.com (Patrick Coskren) Date: Thu Nov 3 14:42:01 2005 Subject: Message-ID: <199905051649.MAA12520@po1.bbn.com> Josh Flowers ecrit: > Is it possible to separate classes by package in PB? I tried > creating subprojects for each package, but when I compiled, the > subprojects appear to get compiled separately, which is problematic > when classes in separate subprojects refer to each other. Is there a > better way to work with java packages in PB, or are subprojects the > accepted way? Request for clarification: are you looking to change anything that affects the code, or just its presentation in PB? If it's the code, then don't worry about it. Just give your .java files the appropriate "package" lines at the top, and the compilation will sort everything out into the right packages when it builds the .jar file. If you're talking about breaking them up in the Project Builder presentation, then I dunno. I kind of miss the "Groups" feature in Metrowerks, myself. Since I think you're talking about the latter, this probably isn't a very helpful message. :-( -Patrick From josh at i33.com Wed May 5 09:59:02 1999 From: josh at i33.com (Josh Flowers) Date: Thu Nov 3 14:42:01 2005 Subject: PB and Java. Message-ID: <199905051659.MAA25928@fnord.i33.com> > Request for clarification: are you looking to change anything that > affects the code, or just its presentation in PB? > Well, a little of both. Changing the presentation is nice, but if you group multiple packages in one project you cannot create two classes that have the same name, but are in different packages. As a note for anyone else looking to do this, do not create your Project as an 'Aggregate', but create it as a 'Java Package'. That was my earlier mistake. From glenn at suntimes.com Wed May 5 15:48:40 1999 From: glenn at suntimes.com (Glenn Carnagey) Date: Thu Nov 3 14:42:01 2005 Subject: WO hosting service Message-ID: <4.2.0.37.19990505174649.00abb670@pop.interaccess.com> Can anyone recommend any hosting services that can handle WebObjects apps? Private replies are probably better. _____________________________________ Glenn Carnagey Webmaster Chicago Sun-Times / Digital Chicago 401 N. Wabash, Rm. 513 Chicago, IL 60611 glenn@suntimes.com, www.suntimes.com 312.321.2603, fax 312.321.2849 _____________________________________ From malcolm at saracen.plsys.co.uk Wed May 5 16:46:56 1999 From: malcolm at saracen.plsys.co.uk (mmalcolm crawford) Date: Thu Nov 3 14:42:01 2005 Subject: WWDC anyone? Message-ID: <199905052346.AAA02417@saracen.plsys.co.uk> David wrote: > But seriously, most of the good WebObjects developers read this list and > I wonder if we are going to have you guys/girls there for support? > Yes, I'll be there. There is going to be some sort of "Yellow Box Birds of a Feather" meeting on Monday night. I'm just arranging details with Scott Anguish at the moment, and they'll be posted on Stepwise as soon as we have final arrangements sorted. Best wishes, mmalc. (Also going to MacWorld NY -- running a Workshop on "Introduction to WO Development" :-) From tom at evatac.com Wed May 5 16:50:23 1999 From: tom at evatac.com (Tom Carstensen) Date: Thu Nov 3 14:42:01 2005 Subject: icons as menu titles Message-ID: <1286174272-2104858@evatac.com> In our carbon/osx version of our apps, we have a script menu that is just a small icon. Under OS X this shows up a garbage text. Is their support for icons as menu titlles under Carbon? We contruct the menu title as described IM Text. Thanks -Tom Carstensen -- Tom Carstensen tom@evatac.com President http://www.evatac.com Evatac Enterprises 877.287.8618 (toll free) Macintosh, Windows & Be Software Development From tom at evatac.com Wed May 5 16:53:03 1999 From: tom at evatac.com (Tom Carstensen) Date: Thu Nov 3 14:42:01 2005 Subject: file format for distribution Message-ID: <1286174112-2114484@evatac.com> We are ready to distribute some of our native carbon apps we built under OS 8.5 using the Carbon SDK. What is the best format so the os x users can upack our archives? .tgz (tar then gzip)? Will using Stuffit to create a .tar, and the using MacGZip to great a .tgz work for osx users? We could zip/tar it up on os x, but it's easier to do it where we develop/compile the carbon app on OS 8.5. Thanks -Tom Carstensen -- Tom Carstensen tom@evatac.com President http://www.evatac.com Evatac Enterprises 877.287.8618 (toll free) Macintosh, Windows & Be Software Development From alexnet at gestalt.com Wed May 5 16:57:13 1999 From: alexnet at gestalt.com (Alex Molochnikov) Date: Thu Nov 3 14:42:01 2005 Subject: Applets in OmniWeb Message-ID: <9905052357.AA01012@gestalt.com> A non-text attachment was scrubbed... Name: not available Type: text/enriched Size: 696 bytes Desc: not available Url : /mailman/archive/macosx-dev/attachments/19990505/839c6e73/attachment.bin From chuck at benatong.com Wed May 5 11:59:11 1999 From: chuck at benatong.com (Charles Bennett) Date: Thu Nov 3 14:42:01 2005 Subject: OBJC to Java question and a few comments. (kinda long) Message-ID: <373094FF.24A8E854@benatong.com> Hi all, I'm "playing" with IB and objc <> java connection s and I seem to have hit a wall that perhaps the list can comment on. For my test I have two objects, one objc and one java, both instantiated in IB the objc object is "wrapped" for exposure to the java object. They each have outlets connected to each other (cross wired) and each expose a method so I can send messages back and forth. I successfully created the bridge with WrapIt to allow the java object to message the objc, with these minor problems. 1: The outlet on the java side must be of type Object or IB can't connect the outlets at run time. I got around this by creating a "set" method. (NeXTSTEP folks.. It's just like the good'ol days :-) ) That is.. assuming that MyObjcObject is a wrapped MyObjcObject.m object MyObjcObject objConnectorOutlet; needs this method public void setObjConnectorOutlet( Object value) { objConnectorOutlet = (MyObjcObject) value; } for the outlet to get connected at run time. Otherwise you take a class cast exception Personally I think this is a bug and that the run time should do the cast and be quiet about it :-) After that, sending messages to the objc object works just fine. String on the java side is NSString* on the objc side.. life is good. However I can't seem to figure out how to go the other way... In my objc object I have the usual object of type id myJavaOutlet and it is connected to the instantiated java object. So.. How do I invoke a method over there? There is no header file to import on the objc side, so it has NO idea what I'm connected to. The "Using java from objective-c" document is less than useless and I can't find any examples that help (We need Librarian back. searching the examples is hard....) It says "In Objctive-C you can use Java classes and interfaces as if they were Objective-C classes and protocols." It then goes on to mumble about perhaps needing NSClassFromString(@"myclass here") this doesn't help because; A: The object is already instantiated. (and must be that way to keep it's outlets connected) B: Since there is no ".h" even if you do create it, you can't send it messages, since objc has no idea what methods exist. The document also mentions something about an explicit mapping in a .jobs file. but I can't find a bit of documentation on how to create a "reverse" .jobs file to map java to objc. Actually I can't even get WrapIt to parse a .java file, It just seems to hang.. Anyone know what I'm doing wrong? Wrong approach? Misc comments on IB and JAVA 1: I seem to have to synchronize my java objects in IB by hand, no amount of dragging and dropping of existing java objects or asking IB to "read from file" seems to help. Is there a way to do this? Trying to keep IB and java objects in "sync" by hand is begging for very hard to find bugs.. Sorry for making a couple of questions into a "rant". Thanks all. chuck From kc at omnigroup.com Wed May 5 17:55:06 1999 From: kc at omnigroup.com (Ken Case) Date: Thu Nov 3 14:42:01 2005 Subject: Applets in OmniWeb Message-ID: <199905060055.RAA06043@ignem.omnigroup.com> > I must be missing something -- perhaps a plug-in? Yes, you're missing the OmniWeb Java plug-in, because it doesn't currently work with X Server: it was written for DR2, and Apple's Java support has changed incompatibly since then. I definitely plan to update the Java plug-in for X Server, but first I need to get OmniGroup Backup out the door. Sorry for the inconvenience! Ken From theroach at csrlink.net Wed May 5 18:01:32 1999 From: theroach at csrlink.net (M. Michael Acosta) Date: Thu Nov 3 14:42:01 2005 Subject: Applets in OmniWeb References: <9905052357.AA01012@gestalt.com> Message-ID: <3730E9E9.D0F96F60@csrlink.net> Alex Molochnikov wrote: > I have a problem displaying Java applet through the OmniWeb > browser in MOSXS. The applet displays (and works) fine > through the applet viewer, but when directed to the browser > page it comes out as an empty rectangular outline with > a cute icon at the bottom. I have the exact same problem. Appletviewer works fine. Java applications I have written work fine on MOSXS, but I can't get Applets to load in OmniWeb. I'd love to hear from someone that knows the simple solution to this problem. Peace, M. Michael Acosta ---------------- Think Different. ---------------- From alexnet at gestalt.com Wed May 5 20:05:20 1999 From: alexnet at gestalt.com (Alex Molochnikov) Date: Thu Nov 3 14:42:01 2005 Subject: Applets in OmniWeb References: <199905060055.RAA06043@ignem.omnigroup.com> Message-ID: <9905060305.AA01198@gestalt.com> A non-text attachment was scrubbed... Name: not available Type: text/enriched Size: 454 bytes Desc: not available Url : /mailman/archive/macosx-dev/attachments/19990505/f92bb282/attachment.bin From sanguish at digifix.com Wed May 5 20:20:25 1999 From: sanguish at digifix.com (Scott Anguish) Date: Thu Nov 3 14:42:01 2005 Subject: file format for distribution In-Reply-To: <1286174112-2114484@evatac.com> References: <1286174112-2114484@evatac.com> Message-ID: <199905060320.XAA25051@digifix.com> Tom Carstensen wrote: > We are ready to distribute some of our native carbon apps we > built under OS 8.5 using the Carbon SDK. > > What is the best format so the os x users can upack our archives? > .tgz (tar then gzip)? > > Will using Stuffit to create a .tar, and the using MacGZip to > great a .tgz work for osx users? The only place that these apps can currently be run on are OS 8.x machines with the Carbon libs installed... They can't run on OS X Server at this point, so... There doesn't seem to be a point to doing anything other than the traditional Mac format of compression (as proprietary as it is). From tom at evatac.com Wed May 5 21:20:51 1999 From: tom at evatac.com (Tom Carstensen) Date: Thu Nov 3 14:42:01 2005 Subject: Carbon API no good w/ Mac OS X?? Message-ID: <1286158045-425759@evatac.com> > > They can't run on OS X Server at this point, so... > > There doesn't seem to be a point to doing anything other than the >traditional Mac format of compression (as proprietary as it is). Wow i'm confused -- i thought that was the whole point of the Carbon SDK... so i can build app that run on OS X. Why wouldn't they run under native OS X? How is anyone supposed to build OS X apps unless they were written in/from the nextstep environment? I thought OS X = Carbon API. Seems silly for apple to bother with the Carbon SDK if there is absolutely no reason to build anything with them. Seems even sillier to release an Mac OS that no exists map app can be ported to. -Tom Carstensen From mer at ma.ultranet.com Wed May 5 19:36:07 1999 From: mer at ma.ultranet.com (Marc Respass) Date: Thu Nov 3 14:42:01 2005 Subject: Keyboard control of UI? Message-ID: <199905060430.AAA16357@ligarius-fe0.ultra.net> Hi, What's the word on keyboard control of the user interface? I have noticed that some MXS apps use keyboard control like highlighting buttons, checkboxes, and radio button and some do not. In IB, I seem to be able to choose if I want to enable keyboard control of UI elements or not. That seems to lend itself to inconsistency - and you know how Mac users hate that ;). So what to do? Should I setup keyboard control of all UI elements and allow tabbing through all controls a la Microsoft Windows or should I make my app more Mac-like and only tab between text fields? Thanks --Marc R From ajetha at tacticalstep.com Wed May 5 21:48:15 1999 From: ajetha at tacticalstep.com (A Jetha) Date: Thu Nov 3 14:42:01 2005 Subject: Carbon API no good w/ Mac OS X?? Message-ID: <7727176C1164.AAA176B@bismuth.tor.sfl.net> > > > > They can't run on OS X Server at this point, so... > > > > There doesn't seem to be a point to doing anything other than the > >traditional Mac format of compression (as proprietary as it is). > > Wow i'm confused -- i thought that was the whole point of the > Carbon SDK... so i can build app that run on OS X. > > Why wouldn't they run under native OS X? How is anyone supposed > to build OS X apps unless they were written in/from the nextstep > environment? I thought OS X = Carbon API. > > Seems silly for apple to bother with the Carbon SDK if there is > absolutely no reason to build anything with them. Seems even > sillier to release an Mac OS that no exists map app can be > ported to. > Carbon is for OS-X --- it is not for OS-X Server. OS-X has not been released yet. (it's easy to confuse the two ...) -- Alykhan Jetha (AJ) TacticalStep Inc., WWW: http://www.tacticalstep.com Email: ajetha@tacticalstep.com Phone: 905.831.2643 ext 1 Fax: 905.837.2546 From hunter at lastonepicked.com Wed May 5 21:47:18 1999 From: hunter at lastonepicked.com (Hunter Hillegas) Date: Thu Nov 3 14:42:01 2005 Subject: Tenon People Message-ID: <19990506044720225.AAA504@nts1.webhostingprovider.com@[207.71.250.180]> Any Tenon people on this list? From td at twics.com Wed May 5 21:54:52 1999 From: td at twics.com (Troy Dawson) Date: Thu Nov 3 14:42:01 2005 Subject: Carbon API no good w/ Mac OS X?? In-Reply-To: <1286158045-425759@evatac.com> Message-ID: On Wed, 5 May 1999, Tom Carstensen wrote: > > They can't run on OS X Server at this point, so... > > > > There doesn't seem to be a point to doing anything other than the > >traditional Mac format of compression (as proprietary as it is). > > Wow i'm confused -- i thought that was the whole point of the > Carbon SDK... so i can build app that run on OS X. Someone correct me if I'm wrong, but I think the Carbon Lib in the SDK is just a cutdown version of InterfaceLib, just so you can check your MacOS apps at linktime for Carbon compatibility. (that's what I'd do if I were in charge of the release schedule). =td= From gideon at csarc.otago.ac.nz Wed May 5 21:59:39 1999 From: gideon at csarc.otago.ac.nz (Gideon.King) Date: Thu Nov 3 14:42:01 2005 Subject: Carbon API no good w/ Mac OS X?? References: <1286158045-425759@evatac.com> Message-ID: <373121BB.1B68C7AF@csarc.otago.ac.nz> You are getting X and X Server confused. X Server in it's current incarnation does not have the Carbon libraries. With MacOS X (and also future versions of Server), you will be able to run both Yellow Box based applicaitons and Carbon applications, taking full advantage of the new OS features. Gideon Tom Carstensen wrote: > > > > They can't run on OS X Server at this point, so... > > > > There doesn't seem to be a point to doing anything other than the > >traditional Mac format of compression (as proprietary as it is). > > Wow i'm confused -- i thought that was the whole point of the > Carbon SDK... so i can build app that run on OS X. > > Why wouldn't they run under native OS X? How is anyone supposed > to build OS X apps unless they were written in/from the nextstep > environment? I thought OS X = Carbon API. > > Seems silly for apple to bother with the Carbon SDK if there is > absolutely no reason to build anything with them. Seems even > sillier to release an Mac OS that no exists map app can be > ported to. > > -Tom Carstensen From rcfa at cubiculum.com Wed May 5 22:53:12 1999 From: rcfa at cubiculum.com (Ronald C.F. Antony) Date: Thu Nov 3 14:42:01 2005 Subject: Keyboard control of UI? In-Reply-To: <199905060430.AAA16357@ligarius-fe0.ultra.net> References: <199905060430.AAA16357@ligarius-fe0.ultra.net> Message-ID: <199905060553.BAA16382@kannix.cubiculum.com> > So what to do? Should I setup keyboard control of all UI elements and > allow tabbing through all controls a la Microsoft Windows or should I make > my app more Mac-like and only tab between text fields? As far as I'm concerned, that's one of the things Windows is good at, even if the implementation sucks: even if a bulldozer just drove over the only mouse in town, you can still get something to print out, edit files, etc. with keyboard control. In old NeXTSTEP and the Mac, you're mostly hosed. OPENSTEP made things a bit better, for the few apps that were fully OpenStep based. I hope that keyboard control is enabled for everything as an alternate way of doing things. (Besides: certain people with motor control disabilities will be thankful, since they often have special keyboards with oversized keys, and can use those to interact with computers...) Ronald ============================================================================== "The reasonable man adapts himself to the world; the unreasonable one persists in trying to adapt the world to himself. Therefore all progress depends on the unreasonable man." G.B. Shaw | rcfa@cubiculum.com | NeXT-mail welcome From kc at omnigroup.com Wed May 5 23:03:48 1999 From: kc at omnigroup.com (Ken Case) Date: Thu Nov 3 14:42:01 2005 Subject: Does OmniWeb expire? Message-ID: <199905060603.XAA27711@ignem.omnigroup.com> > However, the browser's window title on my MOSXS says "Expiring > Beta Release". So, does it or does it not expire? I believe the copy that shipped with X Server still said "Expiring beta release", but if you check the license (select "License..." from the "Help" menu) you'll find it has a free one-user license which expires in 2112. Ken From rknoebl at onevision.de Thu May 6 03:16:21 1999 From: rknoebl at onevision.de (Reiner Knoebl) Date: Thu Nov 3 14:42:01 2005 Subject: C++ Comiler on MacOSX Server Message-ID: <199905061016.LAA03109@dali.onevision.de> A non-text attachment was scrubbed... Name: not available Type: text/enriched Size: 633 bytes Desc: not available Url : /mailman/archive/macosx-dev/attachments/19990506/2fa635f1/attachment.bin From maury at OAAI.COM Thu May 6 07:25:54 1999 From: maury at OAAI.COM (Maury Markowitz) Date: Thu Nov 3 14:42:01 2005 Subject: file format for distribution Message-ID: <199905061422.KAA23402@OAAI.COM> > The only place that these apps can currently be run on are OS 8.x > machines with the Carbon libs installed... > > They can't run on OS X Server at this point, so... > > There doesn't seem to be a point to doing anything other than the > traditional Mac format of compression (as proprietary as it is). Well as a long time Mac user, let me suggest that perhaps now (carbon) is the time to break out of that box. .tgz will work fine on StuffIt as well as other platforms. What's needed is a really good drop-stuffer like ColdCompress for the MacOS. Maury From mike at lorax.com Thu May 6 07:23:05 1999 From: mike at lorax.com (Mike Ferris) Date: Thu Nov 3 14:42:01 2005 Subject: icons as menu titles Message-ID: <199905061518.IAA12717@boom.lorax.com> I don't know how this is done in toolbox programming, but in yellow, you need to use the setImage: API on NSMenuItem to give your menu item an icon (and also do -setTitle:@"" if all you want is the icon). In general, Inside Macintosh is going to be of limited use to yellow developers since the yellow APIs are completely different. Mike Ferris From troy.stephens at page44.com Thu May 6 11:06:10 1999 From: troy.stephens at page44.com (Troy Stephens) Date: Thu Nov 3 14:42:01 2005 Subject: Are the YB dev. tools standard equipment? Message-ID: Apple's vague product literature compels me to ask an ignorant question... Do all the YB/OpenStep development tools (Interface Builder, Project Builder, Yap, frameworks, etc.) come with the standard, off-the-shelf $499 list MacOS X Server distribution? Or are they obtained separately (like it used to be with OpenStep's User vs. Developer CDs)? I feel like an idiot asking this, but I've browsed the specs on Apple's OSX Server user & developer sites, and I even had the OSX Server box in my hand in a store yesterday, and have found no mention of this. (I can understand that they're marketing the Server release primarily to people who want to set up Web servers and do WebObjects stuff, and secondarily to the smaller market of developers trying to get a jump start on the User release, but you'd think either the 4-page PDF "Datasheet" or the 10-page "FAQ" could spare one bullet item to mention the best GUI app development environment this side of a NeXT cube! ;-) Thanks & sorry for the dumb use of bandwidth, Troy Stephens troy.stephens@page44.com From lavoie at cst.ca Thu May 6 11:36:00 1999 From: lavoie at cst.ca (Martin-Gilles) Date: Thu Nov 3 14:42:01 2005 Subject: Are the YB dev. tools standard equipment? Message-ID: <199905061836.OAA32388@plexus.cst.ca> >Apple's vague product literature compels me to ask an ignorant question... > >Do all the YB/OpenStep development tools (Interface Builder, Project >Builder, >Yap, frameworks, etc.) come with the standard, off-the-shelf $499 list MacOS >X Server distribution? Or are they obtained separately (like it used to be >with OpenStep's User vs. Developer CDs)? Standard. MOSXS comes on 4 CDs. One of them is the WebObjects. This one installs add dev kits. Martin-Gilles Lavoie | "No! Try not. Do, or do not. Mac OS programmer, CS&T | There is no try." lavoie@cst.ca | -- Yoda on error handling http://blackhole.cst.ca/ | Y2K statement: hahahaha! From troy.stephens at page44.com Thu May 6 11:36:51 1999 From: troy.stephens at page44.com (Troy Stephens) Date: Thu Nov 3 14:42:01 2005 Subject: Are the YB dev. tools standard equipment? Message-ID: Thanks for the flood of enthusiastic "yes"ses! Clearly there are many NeXT/OpenStep junkies besides myself who are excited about YB development... :) I look forward to learning from & with you all. Troy Stephens troy.stephens@page44.com From raoul at mobopro.com Thu May 6 12:30:21 1999 From: raoul at mobopro.com (Raoul Duke) Date: Thu Nov 3 14:42:01 2005 Subject: Precompiling datedness Message-ID: <3731EDCD.42ADEF3A@mobopro.com> /Local/Users/raoul/Library/Frameworks/AngstOps.framework/Headers/EggOps.h:1: warning: could not use precompiled header '/Local/Users/raoul/Dev/EggOps/EggOps.p', because: /Local/Users/raoul/Library/Frameworks/AngstOps.framework/Headers/EggOps.h:1: warning: 'Operation.h' has different date than in precomp /Local/Users/raoul/Library/Frameworks/AngstOps.framework/Headers/EggOps.h:1: warning: 'Egg/Egg.h' has different date than in precomp /Local/Users/raoul/Library/Frameworks/AngstOps.framework/Headers/EggOps.h:1: warning: 'Egg.h' has different date than in precomp Hi, I've been getting warnings like the ones above when I compile against some of my frameworks. I thought they were due to my using precompiled headers in the frameworks, but I've since turned off the precompiled flags and still get them. I've even "make clean"'d and remade all of the frame works but continue to get them. What do they indicate now and how do I get rid of them? --Raoul From mer at ma.ultranet.com Thu May 6 12:50:30 1999 From: mer at ma.ultranet.com (Marc Respass) Date: Thu Nov 3 14:42:01 2005 Subject: Does OmniWeb expire? Message-ID: <199905061955.PAA22623@ligarius-fe0.ultra.net> >> However, the browser's window title on my MOSXS says "Expiring >> Beta Release". So, does it or does it not expire? > > I believe the copy that shipped with X Server still said "Expiring > beta release", but if you check the license (select "License..." from > the "Help" menu) you'll find it has a free one-user license which > expires in 2112. Expires in 2112, huh? Well, I have to ask - are there some Rush fans at Omnigroup? :). Just a digression, I teach a fibre channel course and the max size of the payload in a fibre channel frame is 2112 bytes. Unfortunately, I hit a generation gap whenever I suggest that Rush fans should easily remembrr that :) --Marc R From gerard at ina.fr Thu May 6 14:02:44 1999 From: gerard at ina.fr (Gerard Iglesias) Date: Thu Nov 3 14:42:01 2005 Subject: Are the YB dev. tools standard equipment? References: Message-ID: <37320373.D69B32C@ina.fr> Troy Stephens wrote: > > Thanks for the flood of enthusiastic "yes"ses! > But, the bundle is limited, I remember that Apple sait somewhere in thei web site that it is a limited offer. I guess that the YB(without WO) will be part of the dev toolkit of MacOS X > Clearly there are many NeXT/OpenStep junkies besides myself who are excited > about YB development... :) As a Mac addict for a very long time, I am happy to see more and more Mac OS developers to be seduced by the YB. > > I look forward to learning from & with you all. > You are welcome :-) Sincerely. Gerard From sanguish at digifix.com Thu May 6 14:03:56 1999 From: sanguish at digifix.com (Scott Anguish) Date: Thu Nov 3 14:42:01 2005 Subject: file format for distribution In-Reply-To: <199905061422.KAA23402@OAAI.COM> References: <199905061422.KAA23402@OAAI.COM> Message-ID: <199905062103.RAA02880@digifix.com> Maury Markowitz wrote: > > The only place that these apps can currently be run on are > OS 8.x > > machines with the Carbon libs installed... > > > > They can't run on OS X Server at this point, so... > > > > There doesn't seem to be a point to doing anything other > than the > > traditional Mac format of compression (as proprietary as it is). > > Well as a long time Mac user, let me suggest that perhaps now > (carbon) is the time to break out of that box. .tgz will work fine > on StuffIt as well as other platforms. What's needed is a really > good drop-stuffer like ColdCompress for the MacOS. Unfortunately, it appears that for Mac OS X (client) they're going to be going with HFS+ as the "standard" file system. That means resource forks and meta-data, which means that without some additional level of encoding (MacBinary lets say) before the tar/compressing, there is STILL going to be problems. ColdCompress will run on Mac OS X.. although what happens as far as the meta data BS is not yet resolved. From sanguish at digifix.com Thu May 6 14:37:31 1999 From: sanguish at digifix.com (Scott Anguish) Date: Thu Nov 3 14:42:01 2005 Subject: Pushing an open compression/archiving format on OSX via dev? In-Reply-To: <199905062103.RAA02880@digifix.com> References: <199905062103.RAA02880@digifix.com> Message-ID: <199905062137.RAA03154@digifix.com> Scott Anguish wrote: > Unfortunately, it appears that for Mac OS X (client) they're going > to be going with HFS+ as the "standard" file system. > > That means resource forks and meta-data, which means that without > some additional level of encoding (MacBinary lets say) before the > tar/compressing, there is STILL going to be problems. I'm not sure if this is appropriate here, Ken can advise where it might be better to discuss this. I'd propose a setting up a working group to formalize our own open standard for this, using existing tools and protocols in a combination such that this can be accomplished easily... I'm thinking of some sort of a combination of - AppleSingle/AppleDouble or MacBinary for maintaining resources/metadata - gnutar for archiving - gzip or bzip for compression.. Possibly including a small header on the front of each file (easily stripped off using pretty much any shell tools) that could contain additional information about the archive (what is in it for example, creating app, splash-screen, etc) so that the actual archive doesn't need to be decompressed/untarred to get this information. If anyone is interested, we can setup a small discussion group post WWDC.... From cmh at greendragon.com Thu May 6 14:52:10 1999 From: cmh at greendragon.com (Chris Hanson) Date: Thu Nov 3 14:42:01 2005 Subject: file format for distribution In-Reply-To: <199905062103.RAA02880@digifix.com> Message-ID: At 4:22 PM -0500 5/6/99, Scott Anguish wrote: > Unfortunately, it appears that for Mac OS X (client) they're going >to be going with HFS+ as the "standard" file system. > > That means resource forks and meta-data, which means that without >some additional level of encoding (MacBinary lets say) before the >tar/compressing, there is STILL going to be problems. HFS+ is by no means the only filesystem with support for metadata and additional forks. NTFS is even more flexible than HFS+. Most Unix flavors support real access control lists, with varying levels of complexity. I think SGI's XFS supports tagged metadata. From what I've heard about the BeOS filesystem it seems similar to NTFS. A couple of years ago some BeOS developers were talking about creating a new compression and archive format which would allow for quick browsing like StuffIt and Zip , large (64-bit sized) files, *and* the ability to avoid metadata hacks like MacBinary or AppleDouble-style encoding. I wonder if it went anywhere; it'd probably be useful on all of the above systems. If anyone wants to look for it the archives of the BeDevTalk list would be a good starting point. From krw at nobugz.com Thu May 6 15:06:53 1999 From: krw at nobugz.com (Randy Wigginton) Date: Thu Nov 3 14:42:01 2005 Subject: $20 for David? In-Reply-To: <199905061518.IAA12717@boom.lorax.com> Message-ID: OK, here's another question that is most likely really really simple, but its driving me nuts... How do I get the classes to show up in PB in alphabetical order? I've tried the obvious suspects. BTW, is Purge Indexes really supposed to shut down Project Builder? OK, Dave, time's a tickin... Its 3:07 here in Cupertino... --randy From rcfa at cubiculum.com Thu May 6 14:42:12 1999 From: rcfa at cubiculum.com (Ronald C.F. Antony) Date: Thu Nov 3 14:42:01 2005 Subject: file format for distribution In-Reply-To: <199905062103.RAA02880@digifix.com> References: <199905062103.RAA02880@digifix.com> Message-ID: <199905062142.RAA19405@kannix.cubiculum.com> > Unfortunately, it appears that for Mac OS X (client) they're going > to be going with HFS+ as the "standard" file system. I thought that only legacy, i.e. transparent-blue-box apps will make use of them, and that Carbon/CoreFoundation based apps will use the bundle scheme. > That means resource forks and meta-data, which means that without > some additional level of encoding (MacBinary lets say) before the > tar/compressing, there is STILL going to be problems. Unless you have new information that contradicts the above assumption, that would be easy: only files with non-empty forks would cause a problem, everything else would simply be automatically handled like a flat file and would not transmit an empty fork structure, with all the problems that creates. Ronald ============================================================================== "The reasonable man adapts himself to the world; the unreasonable one persists in trying to adapt the world to himself. Therefore all progress depends on the unreasonable man." G.B. Shaw | rcfa@cubiculum.com | NeXT-mail welcome From toon at omnigroup.com Fri May 7 16:25:30 1999 From: toon at omnigroup.com (Greg Titus) Date: Thu Nov 3 14:42:01 2005 Subject: $20 for David? Message-ID: <199905062326.QAA03580@ignem.omnigroup.com> Randy Wigginton wrote: > How do I get the classes to show up in PB in alphabetical order? You can either Control-drag the items in the list into whatever order you want, or you can go here: http://www.omnigroup.com/DeveloperResources/PBBundles/OmniPBSortedComponents.html ... and get a project builder bundle that gets loaded in and will keep all your files in alphabetical order automatically. Hope this helps, --Greg ------------ Greg Titus Omni Development Ic. greg@omnigroup.com From sanguish at digifix.com Thu May 6 20:13:18 1999 From: sanguish at digifix.com (Scott Anguish) Date: Thu Nov 3 14:42:01 2005 Subject: file format for distribution In-Reply-To: References: Message-ID: <199905070313.XAA04693@digifix.com> Chris Hanson wrote: > At 4:22 PM -0500 5/6/99, Scott Anguish wrote: > >Unfortunately, it appears that for Mac OS X (client) they're going > >to be going with HFS+ as the "standard" file system. > > > >That means resource forks and meta-data, which means that without > >some additional level of encoding (MacBinary lets say) before the > >tar/compressing, there is STILL going to be problems. > > HFS+ is by no means the only filesystem with support for metadata and > additional forks. NTFS is even more flexible than HFS+. But have you EVER encountered a forked file on NT? They all just can be archived and restored by basic stuff... > A couple of years ago some BeOS developers were talking about creating a > new compression and archive format which would allow for quick browsing > like StuffIt and Zip , large (64-bit sized) files, *and* the ability to > avoid metadata hacks like MacBinary or AppleDouble-style encoding. MacBinary and AppleSingle aren't really hacks to store the info, they're just standard formats for represending them as flat files.. (Apple Double splits it across multiple files/directories... but maybe thats the best way to handle it.) From louis.pelletier at altavista.net Thu May 6 20:46:14 1999 From: louis.pelletier at altavista.net (Louis Pelletier) Date: Thu Nov 3 14:42:01 2005 Subject: file format for distribution Message-ID: <199905070344.UAA17513@ignem.omnigroup.com> >De?: Scott Anguish >? : Multiple recipients of list >Objet?: Re: file format for distribution >Date?: Jeud 06 mai 1999 23:25 > > Chris Hanson wrote: >> At 4:22 PM -0500 5/6/99, Scott Anguish wrote: >> >Unfortunately, it appears that for Mac OS X (client) they're going >> >to be going with HFS+ as the "standard" file system. >> > >> >That means resource forks and meta-data, which means that without >> >some additional level of encoding (MacBinary lets say) before the >> >tar/compressing, there is STILL going to be problems. >> >> HFS+ is by no means the only filesystem with support for metadata and >> additional forks. NTFS is even more flexible than HFS+. > > But have you EVER encountered a forked file on NT? Correct me if I'm wrong, but I believe all security info is stored in forks.. And I think the sluggish AppleTalk server that ships with NT Server uses NTFS forks to store HFS ressources... [...] -- Louis Pelletier, Quebec From sanguish at digifix.com Thu May 6 21:17:04 1999 From: sanguish at digifix.com (Scott Anguish) Date: Thu Nov 3 14:42:01 2005 Subject: file format for distribution In-Reply-To: <199905070344.UAA17513@ignem.omnigroup.com> References: <199905070344.UAA17513@ignem.omnigroup.com> Message-ID: <199905070417.AAA04841@digifix.com> Louis Pelletier wrote: > > But have you EVER encountered a forked file on NT? > > Correct me if I'm wrong, but I believe all security info is stored in > forks= .. But this is volatile information that wouldn't be transported to other machines correct? > And I think the sluggish AppleTalk server that ships with NT Server uses > NTFS forks to store HFS ressources... I wouldnt' consider that native use of that though.. From louis.pelletier at altavista.net Fri May 7 00:58:13 1999 From: louis.pelletier at altavista.net (Louis Pelletier) Date: Thu Nov 3 14:42:01 2005 Subject: file format for distribution Message-ID: <199905070756.AAA03509@ignem.omnigroup.com> >De?: Scott Anguish >? : Multiple recipients of list >Objet?: Re: Re : file format for distribution >Date?: Vend 07 mai 1999 00:29 > > Louis Pelletier wrote: >> > But have you EVER encountered a forked file on NT? >> >> Correct me if I'm wrong, but I believe all security info is stored in >> forks= .. > > But this is volatile information that wouldn't be transported to > other machines correct? Humm correct... Does NTFS security info get transported to other FS? I guess so, but that would mean that either NT knows how to read its own forked security to write it, let's say on ufs; or either Unix, with an NTFS driver reads the forked security on NTFS... But any way, it means that NTFS uses uncompatible forked FS... to use the argument many use against HFS/HFS+... Of course, losing half of the data is not like losing security info...! -- Louis Pelletier, Quebec From uscheidegger at macropolis.ch Fri May 7 05:35:11 1999 From: uscheidegger at macropolis.ch (Ulrich Scheidegger) Date: Thu Nov 3 14:42:01 2005 Subject: HTMLEditor - Series cancelled Message-ID: <199905071214.OAA18812@mailout.access.ch> Sorry to hear about your trouble with Apple! Well, thanks anyway for publishing an article series for novices like me. The TravelAdvisory etc. projects from Apple did not answer many questions. It would be nice though, if one could download all 7 articles including the graphics etc. in one folder, so that one could work with them offline. Thanks again for your effort. Will From jean-michel.cazaux at finindev.com Fri May 7 06:11:23 1999 From: jean-michel.cazaux at finindev.com (Jean-Michel Cazaux) Date: Thu Nov 3 14:42:01 2005 Subject: Adding behaviour when (subclass of) button is clicked... Message-ID: <01BE989B.E1084140@JEAN-MICHEL> We've got a subclass of NSButton and we'd like to add behaviour when the button is clicked... (go to the next key view). Any idea on what is the "right" method to override ??? Thanks in advance for any help... __________________________________________________________________________________________ Jean-Michel CAZAUX FININDEV, Conseil en Finances pour les Collectivit?s Locales. 204 Rue Michel Teule - ZAC d'Alco 34080 Montpellier - FRANCE. T?l. +33 (0)4 67 63 66 25 - Fax +33 (0)4 67 63 35 45 e-mail: jean-michel.cazaux@finindev.com From alberto at inis.de Fri May 7 07:36:32 1999 From: alberto at inis.de (Alberto Araoz) Date: Thu Nov 3 14:42:01 2005 Subject: Adding behaviour when (subclass of) button is clicked... References: <01BE989B.E1084140@JEAN-MICHEL> Message-ID: <199905071436.PAA00972@inis-5.inis.de> mouseDown (inherited from NSControl) From cmh at greendragon.com Fri May 7 06:39:11 1999 From: cmh at greendragon.com (Chris Hanson) Date: Thu Nov 3 14:42:01 2005 Subject: file format for distribution In-Reply-To: <199905070313.XAA04693@digifix.com> Message-ID: Before arbitrarily deciding to base a new distribution format on gnutar and gzip, why not take the time to understand why Macintosh users use StuffIt? It's not just that Expander and DropStuff are fully-functional shareware; if that were all there was to it, Aladdin wouldn't be around now. StuffIt has a great interface and its file format is likely designed around that. I don't believe such an interface can be built on top of compressed tar files. Oh, I'm sure it *technically* could, but it would be much slower at manipulating files within archives. Perhaps Zip is a better starting point... From raoul at mobopro.com Fri May 7 08:46:55 1999 From: raoul at mobopro.com (Raoul Duke) Date: Thu Nov 3 14:42:01 2005 Subject: (SUMMARY) Precompiling datedness References: <3731EDCD.42ADEF3A@mobopro.com> Message-ID: <37330AEF.B4DCDA58@mobopro.com> I'm not sure how I overlooked it, but there was a .p file in the project directory of one of my fundamental frameworks. Apparently, make clean doesn't wipe these out. Best, Raoul From ayoungs at rochester.rr.com Fri May 7 09:26:51 1999 From: ayoungs at rochester.rr.com (Andy Youngs) Date: Thu Nov 3 14:42:01 2005 Subject: School project with ipc.h Message-ID: I've had to resort back to Solaris box to do a school project. I'm curious tho why the following weren't in the include sys #include #include These files I needed aren't available. They deal with shared memory and process communication. Is it that these are specific to Solaris? or would a Linux box have these? Regards Andy From cpry at enc.org Fri May 7 10:33:16 1999 From: cpry at enc.org (chad pry) Date: Thu Nov 3 14:42:01 2005 Subject: compiling applications Message-ID: <000601be98af$b09a6600$7b619280@enc.org> i am new to the macOS X arena. Does macOS X come w/a precompiled make/gcc libraries under a normal install? I was not able to find the make or gcc libraries in my path or after doing a system-wide find. Could someone help me in this area. thanks, Chad W. Pry Systems Developer Eisenhower National Clearinghouse http://www.enc.org/ mailto:cpry@enc.org phone 614.292.9996 fax 614.292.2066 From kc at omnigroup.com Fri May 7 13:14:59 1999 From: kc at omnigroup.com (Ken Case) Date: Thu Nov 3 14:42:01 2005 Subject: School project with ipc.h Message-ID: <199905072015.NAA09168@ignem.omnigroup.com> > I've had to resort back to Solaris box to do a school project. > I'm curious tho why the following weren't in the include sys > #include > #include Those are System V IPC mechanisms. (BSD invented the socket mechanism for IPC instead.) BSD's sockets have now been included on System V platforms (and most non-UNIX platforms), but not (generally) vice versa. > Is it that these are specific to Solaris? They're definitely not specific to Solaris: they were on AT&T 3B2 systems (running System V) long before Sun switched from a BSD-based UNIX (SunOS 4.x and earlier) to a System V-based UNIX (Solaris 2.x and later). > or would a Linux box have these? Since Linux is based on System V it might have them, but I don't know for certain. Ken From kc at omnigroup.com Fri May 7 13:17:42 1999 From: kc at omnigroup.com (Ken Case) Date: Thu Nov 3 14:42:01 2005 Subject: compiling applications Message-ID: <199905072017.NAA09226@ignem.omnigroup.com> > Does macOS X come w/a precompiled make/gcc libraries under a > normal install? They're in the developer packages on the WebObjects CD-ROM. (You don't actually have to install WebObjects, you can just install the DeveloperTools and DeveloperLibs packages.) Ken From norbert.heger at obdev.at Fri May 7 14:07:51 1999 From: norbert.heger at obdev.at (Norbert Heger) Date: Thu Nov 3 14:42:01 2005 Subject: Adding behaviour when (subclass of) button is clicked... References: <01BE989B.E1084140@JEAN-MICHEL> Message-ID: <9905072107.AA03426@cray> Jean-Michel Cazaux asked: > We've got a subclass of NSButton and we'd like to add behaviour > when the button is clicked... (go to the next key view). > Any idea on what is the "right" method to override ??? Alberto Araoz replied: > mouseDown (inherited from NSControl) I'd suggest to use 'sendAction:to:' instead, since this method is only invoked when the button is 'really' clicked. In contrast to 'mouseDown:' it is not called when you cancel the click releasing the mouse outside the NSButton frame. Hope this helps, Norbert _____________________________ _________________________________ O B J E C T I V E D E V E L O P M E N T Norbert Heger OpenStep MacOS-X EOF WebObjects norbert.heger@obdev.at Software Development +43-664-2811777 www.obdev.at From bbum at codefab.com Fri May 7 14:25:28 1999 From: bbum at codefab.com (Bill Bumgarner) Date: Thu Nov 3 14:42:02 2005 Subject: Adding behaviour when (subclass of) button is clicked... References: <9905072107.AA03426@cray> Message-ID: <37335A48.B5FB50A3@mail.codefab.com> This really doesn't sound like the right behaviour to be implemented in an NSButton. NSButtons represent the concept of an object that tracks the mouse and sends its target an action if and only if the user successfully clicked the button. They are not intended to know anything about view hierarchies, display logic, navigation logic, etc... The target of the button should be the object that takes the user to the next key view. Typically, the target of the button would be some kind of controller object. I would highly recommend reading "Objective-C and Object Oriented Programming" followed by whatever discussions you can find on Model-View-Controller application design and construction. Following an MVC paradigm will lead to a much more consistent application design and behvaiour with a codebase that can migrate between different environments with considerably less effort. b.bum Norbert Heger wrote: > Jean-Michel Cazaux asked: > > We've got a subclass of NSButton and we'd like to add behaviour > > when the button is clicked... (go to the next key view). > > Any idea on what is the "right" method to override ??? > > Alberto Araoz replied: > > mouseDown (inherited from NSControl) > > I'd suggest to use 'sendAction:to:' instead, since this method is only > invoked when the button is 'really' clicked. In contrast to 'mouseDown:' > it is not called when you cancel the click releasing the mouse outside > the NSButton frame. > > Hope this helps, > Norbert > > _____________________________ _________________________________ > O B J E C T I V E D E V E L O P M E N T > > Norbert Heger OpenStep MacOS-X EOF WebObjects > norbert.heger@obdev.at Software Development > +43-664-2811777 www.obdev.at From alexnet at gestalt.com Fri May 7 14:28:06 1999 From: alexnet at gestalt.com (Alex Molochnikov) Date: Thu Nov 3 14:42:02 2005 Subject: Adding behaviour when (subclass of) button is clicked... Message-ID: <9905072128.AA00816@gestalt.com> A non-text attachment was scrubbed... Name: not available Type: text/enriched Size: 1153 bytes Desc: not available Url : /mailman/archive/macosx-dev/attachments/19990507/1018649a/attachment.bin From ngustilo at earthlink.net Fri May 7 15:20:39 1999 From: ngustilo at earthlink.net (Nicholas A. Gustilo) Date: Thu Nov 3 14:42:02 2005 Subject: NSStatusItems and Menus and Submenus Message-ID: <199905072222.PAA20112@grebe.prod.itd.earthlink.net> Hi, I'm having a problem using NSStatusItem with a submenu. I create the NSStatusItem and set a menu that has submenus. When I run my app the menus and submenus in the NSStatusItem show up. However, if I'm in a different app and click on the NSStatusItem only the menu items show up (the submenu items don't show up). Anyone have any ideas why this would be? I'm running Mac X server on a G3 Powermac. Any help is appreciated. Nick Gustilo From rlove at neosoft.com Fri May 7 15:43:05 1999 From: rlove at neosoft.com (Robert B. Love) Date: Thu Nov 3 14:42:02 2005 Subject: file format for distribution In-Reply-To: References: Message-ID: <199905072243.RAA00987@neosoft.com> You wrote: > Before arbitrarily deciding to base a new distribution format on > gnutar and gzip, why not take the time to understand why Macintosh > users use StuffIt? First lets ask how good is StuffIt. The reason I want to know is when I bought version 5 recently it said "Now with 20% better compression". Doesn't sound like it was all that good at its main function. Maybe its better now. And is it Opensource? Can we base a tool we all work on without source? I bet if the source had been available the compression would have been better long ago. ---------------------------------------------------------------- Bob Love MIME & NeXT Mail OK rlove@neosoft.com ---------------------------------------------------------------- From epeyton at epicware.com Fri May 7 16:18:57 1999 From: epeyton at epicware.com (Eric S. Peyton) Date: Thu Nov 3 14:42:02 2005 Subject: NSStatusItems and Menus and Submenus In-Reply-To: <199905072222.PAA20112@grebe.prod.itd.earthlink.net> Message-ID: Just for the record I have also had this problem (with Fire.app) and could not find a work around ... Eric Eric Peyton epeyton@epicware.com Software and Source for Mac OS X Server On Fri, 7 May 1999, Nicholas A. Gustilo wrote: > Hi, I'm having a problem using NSStatusItem with a submenu. I create the > NSStatusItem and set a menu that has submenus. When I run my app the menus > and submenus in the NSStatusItem show up. However, if I'm in a different > app and click on the NSStatusItem only the menu items show up (the submenu > items don't show up). Anyone have any ideas why this would be? I'm running > Mac X server on a G3 Powermac. Any help is appreciated. Nick Gustilo > > From raoul at mobopro.com Fri May 7 16:19:42 1999 From: raoul at mobopro.com (Raoul Duke) Date: Thu Nov 3 14:42:02 2005 Subject: Not a concrete instance?? Message-ID: <3733750E.6D9152F2@mobopro.com> Hi, I have a question so basic that I'm lost. When I invoke this line: Column *col = [[Column alloc] initWithCapacity:rows]; I get this runtime error: (gdb) May 07 16:11:12 Hmm[1449] An uncaught exception was raised May 07 16:11:12 Hmm[1449] *** initialization method -initWithCapacity: cannot be sent to an abstract object of class Column: Create a concrete instance! May 07 16:11:12 Hmm[1449] *** Uncaught exception: *** initialization method -initWithCapacity: cannot be sent to an abstract object of class Column: Create a concrete instance! stack: 0x4257d4e4 0x42586680 0x42583c58 0x42581a5c 0x425819c4 0x4257f6b8 0x42569368 0x3500 0x3914 0x36e4 0x4337b9b4 0x433047b8 0x433679f0 0x43366de0 0x4336578c 0x43364970 0x43397920 0x43503b58 0x7c14 0x33f0 0x32b0 Program exited with code 0377. No stack. The Column class is listed completely below. As you can see, it's extremely simple. Interestingly, when I invoke this line with an NSMutableArray instead of Column, it works fine. Thanks as always, Raoul #import @interface Column:NSMutableArray { float width; } - initWithCapacity:(unsigned int)numRows; - (void)setWidth:(float)aWidth; - (float)width; @end #import "Column.h" @implementation Column #define DEFAULT_WIDTH 50 - initWithCapacity:(unsigned int)numRows { [super initWithCapacity:numRows]; width = DEFAULT_WIDTH; return self; } - (void)setWidth:(float)aWidth { width = aWidth; } - (float)width { return width; } @end From cwhite at comnetix.com Fri May 7 17:54:33 1999 From: cwhite at comnetix.com (Craig White) Date: Thu Nov 3 14:42:02 2005 Subject: Crash in NSCursor Message-ID: <9905072054.AA335782@deli> We've been getting periodic crashes in our application in NSCursor. Has anyone else experienced this on OE4.2? We use a new category method called +displayWaitCursorUntilIdle that displays a busy cursor and does a delayed method call to display the arrow cursor again (see below). Not the best strategy, but I can't see why it would crash... Any ideas? Craig May 07 16:44:16 RICI[280] Caught WIN32 exception 0xc0000005: "Access Violation on read at address 0xd1f000" Attempting stack frame backtrace: function(0x12fa44, 0x25ef008, 0x000010, 0x000010) - [NSBitmapImageRep _getEightBitRGBMeshedBitmap:(char *) rowBytes:(int)64 extraSample:(char)\000 reverseScanLines:(char)\000 removeAlpha:(char)\000] function(0x6a8d80, 0x40e00000, 0x41000000, 0x6a8dc0) - [NSCursor _setWindowsCursor] - [NSCursor set] function(0x6afdb0, 0x43f00000, 0x437d0000, 0x000200) function(0x127052c, 0x000200, 0x000000, 0x1701e0) function(0x127052c, 0x000200, 0x000000, 0x1701e0) function(0x127052c, 0x000200, 0x000000, 0x1701e0) function(0x505094, 0x000361, 0x0000e3, 0x12fcfc) static NSCursor *busyCursor = nil; @implementation NSCursor (Spin) + (void)displayWaitCursorUntilIdle { [self displayWaitCursor]; [self performSelector:@selector( displayArrowCursor ) withObject:nil afterDelay:0]; } + (void)displayWaitCursor { if( !busyCursor ) { NSImage *busyImage; int *ptr; ptr = LoadCursor(NULL,IDC_WAIT); busyImage = [[NSImage alloc] initWithIconHandle:ptr]; busyCursor = [[self alloc] initWithImage:busyImage hotSpot:NSZeroPoint]; } [busyCursor set]; } + (void)displayArrowCursor { [[self arrowCursor] set]; } @end From richs at win.bright.net Sat May 8 09:38:14 1999 From: richs at win.bright.net (Rich Schroedel) Date: Thu Nov 3 14:42:02 2005 Subject: Yellow Box Supplemental Message-ID: <37346874.A5FFDB4D@win.bright.net> After downloading "PROGRAMMING YELLOW BOX Rhapsody Developer Release Supplemental", one of the pdf files, Chapter 1, Frameworks, was unreadable. Does anyone know where I could find a good copy? It seems to me I've seen mention of this before, either on the dev list or the talk list. I tried to find the reference in Omni's archive search, but it seems to be down. -- Rich Schroedel richs@win.bright.net From cmh at greendragon.com Sat May 8 12:09:44 1999 From: cmh at greendragon.com (Chris Hanson) Date: Thu Nov 3 14:42:02 2005 Subject: file format for distribution In-Reply-To: <199905072243.RAA00987@neosoft.com> References: Message-ID: I don't think you understand my point. Some people want to design a competitor to StuffIt. Understanding *why* people use StuffIt in the first place is a necessary part of designing a competing product -- no matter whether that product is commercial or free or public domain or Open Source -- because otherwise consumers are likely to continue using StuffIt as long as it's available on their main platform. By your statements below, it might be safe to conclude that StuffIt users aren't basing their decision to use it solely on its compression abilities. My guess is that its compression is good enough for most users, and that what causes people to buy the commercial StuffIt Deluxe package is its interface and/or the system enhancements that it includes. I know that's why I bought it; I like being able to open up an archive, rename files and folders within it, move things around inside of it, etc. These are all things which will be hard to implement and time-consuming for users in any competitor based on tar & gzip because of how those tools work. At 5:43 PM -0500 5/7/99, Robert B. Love wrote: >You wrote: >> Before arbitrarily deciding to base a new distribution format on >> gnutar and gzip, why not take the time to understand why Macintosh >> users use StuffIt? > >First lets ask how good is StuffIt. The reason I want to know >is when I bought version 5 recently it said "Now with 20% better >compression". Doesn't sound like it was all that good at its >main function. Maybe its better now. > > >And is it Opensource? Can we base a tool we all work on without >source? I bet if the source had been available the compression >would have been better long ago. From Philippe.Robert at uptime.ch Sat May 8 03:56:37 1999 From: Philippe.Robert at uptime.ch (Philippe C.D. Robert) Date: Thu Nov 3 14:42:02 2005 Subject: Not a concrete instance?? In-Reply-To: <3733750E.6D9152F2@mobopro.com> References: <3733750E.6D9152F2@mobopro.com> Message-ID: <9905082000.AA00562@uptime.ch> You wrote: > Hi, > > I have a question so basic that I'm lost. When I invoke this line: > > Column *col = [[Column alloc] initWithCapacity:rows]; > > I get this runtime error: > [snip] You can't do this, collection classes are class clusters! Instead why not have a NSMutableArray as an ivar in your Column class?! Such as: @interface Column:NSObject { NSMutableArray array; float width; } andsoonandsoon...;-) Then you can do a array = [[NSMutableArray alloc] initWithCapacity:bla]; in init of Column. sweet dreams, Phil --- Philippe C.D. Robert Uptime Object Factory Inc http://www.nice.ch/~phip Software Development http://www.projectcenter.ch OpenStep/Unix From kelleysoft at pressenter.com Sat May 8 18:53:54 1999 From: kelleysoft at pressenter.com (Greg Betzel) Date: Thu Nov 3 14:42:02 2005 Subject: Arranging nibs in PB?? Message-ID: Has anyone noticed that arranging nib files in the MacOS X PB browser doesn't "take"? You can control-drag the list items to where you want, but they revert back to thier old positions when you browse someplace else & come back. This problem doesn't exist with the .m & .h files, and it didn't exist in the DR1/DR2 PB. Does anyone have any guesses why or ideas on a workaround?? I have a project that has a LOT of nibs, and having them all jumbled together doesn't cut it at all. Omni's alphabetizer bundle would help, I suppose, but I'd like to have things work right if possible. Thanks in advance. Greg Betzel From karmat at matthias.org Sat May 8 19:37:37 1999 From: karmat at matthias.org (Karl N. Matthias) Date: Thu Nov 3 14:42:02 2005 Subject: School project with ipc.h In-Reply-To: <199905072015.NAA09168@ignem.omnigroup.com> Message-ID: See below On Fri, 7 May 1999, Ken Case wrote: > > I've had to resort back to Solaris box to do a school project. > > I'm curious tho why the following weren't in the include sys > > #include > > #include > > Those are System V IPC mechanisms. (BSD invented the socket > mechanism for IPC instead.) BSD's sockets have now been included on > System V platforms (and most non-UNIX platforms), but not (generally) > vice versa. Most of the modern BSD-based systems(OpenBSD, FreeBSD, NetBSD, Digital UNIX) do, in fact, include SysV IPC. Haven't bothered to look for it on MOSX but I would have expected it to be there. SysV IPC has pretty much become a 'standard' part of most Unixen. It is also part of the X/Open XPG4 and XPG4-UNIX specifications. Is this yet another thing Apple left out? > > > Is it that these are specific to Solaris? > > They're definitely not specific to Solaris: they were on AT&T 3B2 > systems (running System V) long before Sun switched from a BSD-based > UNIX (SunOS 4.x and earlier) to a System V-based UNIX (Solaris 2.x > and later). > > > or would a Linux box have these? > > Since Linux is based on System V it might have them, but I don't > know for certain. > I don't know of a Linux distrib which doesn't have them. Karl From eric at EMIEng.com Sat May 8 19:42:51 1999 From: eric at EMIEng.com (Eric Marshall) Date: Thu Nov 3 14:42:02 2005 Subject: making a window key when not active application Message-ID: <199905090242.WAA29100@bobo.shirenet.com> Using NSWindow's makeKeyAndOrderFront: method, I'm trying to make a window the key window even though the application is not the active one, but it's not working for me. The documentation for makeKeyAndOrderFront: doesn't mention a prerequisite that the application must be active for the method to function as advertised. I've seen a couple of other applications recently that demonstrate the same behavior as my test app, but Workspace Manager seems to be able to do it. Any tips on this? Thanks in advance. From Izidor.Jerebic at select-tech.si Sun May 9 07:15:49 1999 From: Izidor.Jerebic at select-tech.si (Izidor Jerebic) Date: Thu Nov 3 14:42:02 2005 Subject: non-rectangular clippaths for EPS images do not work on screen Message-ID: <199905091414.QAA26205@stsrv.select-tech.si> Hello! The title says it all... If an EPS image is drawn (via NSImage/NSEPSImageRep) after a non-rectangular clip path has been set, the image on screen clips to clip path's bounding box... The print output is clipped OK, the EPS output is clipped OK, just the screen output is not... If we combine the clip path and the EPS image in another EPS image, the resulting image displays correctly (namely, original image clipped according to the path). If the image is TIFF, it clips correctly... We work on OpenStep 4.2 and MacOS-X-Server... Recombining clip-path and image in new EPS image is too slow and unreliable... Any workarounds, hints ? Regards, Izidor Jerebic SELECT Technology From kc at omnigroup.com Sun May 9 15:37:39 1999 From: kc at omnigroup.com (Ken Case) Date: Thu Nov 3 14:42:02 2005 Subject: School project with ipc.h In-Reply-To: <199905072015.NAA09168@ignem.omnigroup.com> Message-ID: <199905092237.PAA18483@ignem.omnigroup.com> > Most of the modern BSD-based systems(OpenBSD, FreeBSD, NetBSD, > Digital UNIX) do, in fact, include SysV IPC. [...] The Mach kernel was designed to support BSD 4.3, which predated System V IPC in BSD. To my knowledge, Mach 2.x has no support for true shared memory between tasks (it has copy-on-write memory and some hacks for drivers), so adding support for System V IPC isn't just a matter of porting the latest BSD 4.4 code, it needs some changes to the Mach kernel as well. (Well, that's not strictly true: you could emulate the shared memory facilities, and in fact there's a library out there somewhere which does this, but it's not nearly as efficient as the real thing.) > Is this yet another thing Apple left out? I'd expect Apple to wait until they switch to the new Mach 3.x-based kernel in OS X before making any significant changes or additions to Mach. Ken From gerard at ina.fr Mon May 10 01:18:27 1999 From: gerard at ina.fr (Gerard Iglesias) Date: Thu Nov 3 14:42:02 2005 Subject: non-rectangular clippaths for EPS images do not work on screen References: <199905091414.QAA26205@stsrv.select-tech.si> Message-ID: <37369654.B94BD0AE@ina.fr> Izidor Jerebic wrote: > If an EPS image is drawn (via NSImage/NSEPSImageRep) after a > non-rectangular clip path has been set, the image on screen clips to > clip path's bounding box... > The print output is clipped OK, the EPS output is clipped OK, just > the screen output is not... > If we combine the clip path and the EPS image in another EPS image, > the resulting image displays correctly (namely, original image > clipped according to the path). > > If the image is TIFF, it clips correctly... > > We work on OpenStep 4.2 and MacOS-X-Server... > > Recombining clip-path and image in new EPS image is too slow and > unreliable... > > Any workarounds, hints ? > I think that the problem come from NSImage compositing method. For performance reason, NSImage compsite method make the minimum work, you can't take in consideration the current space transformation, excepted the current origin. I think that you have to try the function NSDrawBitmap or the draw/drawAtPoint NSImageRep method. Obviously these method can't handle compositing. I think that those method are use when printing image. Sincerely. -- G?rard Iglesias Ph.D. in Computer Graphic gerard.iglesias@ina.fr Researcher/Project Manager http://www.mediaport.net/People/Gerard.Iglesias From Izidor.Jerebic at select-tech.si Mon May 10 01:28:52 1999 From: Izidor.Jerebic at select-tech.si (Izidor Jerebic) Date: Thu Nov 3 14:42:02 2005 Subject: non-rectangular clippaths for EPS images do not work on screen In-Reply-To: <37369654.B94BD0AE@ina.fr> References: <199905091414.QAA26205@stsrv.select-tech.si> <37369654.B94BD0AE@ina.fr> Message-ID: <199905100827.KAA18538@stsrv.select-tech.si> >I think that the problem come from NSImage compositing method. For >performance reason, NSImage compsite method make the minimum work, >you >can't take in consideration the current space transformation, >excepted >the current origin. >I think that you have to try the function NSDrawBitmap or the >draw/drawAtPoint NSImageRep method. Obviously these method can't >handle >compositing. I think that those method are use when printing image. The image is drawn as follows: [[image bestRepresentationForDevice:nil] draw] ; If this image was created from EPS data, then on screen it is clipped to bounding box of clip path, not to clip path !!! I guess this has something to do with internal image cacheing/display mechanism (bug?) of NSEPSImageRep and window server, because with TIFF images everything works OK. Regards, Izidor Jerebic SELECT Technology From tbumgarner at pobox.com Mon May 10 08:19:34 1999 From: tbumgarner at pobox.com (Tim Bumgarner) Date: Thu Nov 3 14:42:02 2005 Subject: TIFF images Message-ID: <1285801276-212351166@macitools.symantec.com> What is the best way to prepare TIFF images to be used as icons, especially in regards to the alpha channels? I've tried PhotoShop 5 and created an icon with either a straight alpha channel or the quick mask feature and saved the icon as a TIFF. When I use the icon in a MOXS app, I get the following warning in the log window: Warning: TIFF image with unknown extra samples assumed to have unassociated alpha. RGB values have been premultiplied The icon works fine, and has the correct Alpha (or mask). Anyone have any ideas? twb From gerard at ina.fr Mon May 10 09:29:28 1999 From: gerard at ina.fr (Gerard Iglesias) Date: Thu Nov 3 14:42:02 2005 Subject: TIFF images References: <1285801276-212351166@macitools.symantec.com> Message-ID: <37370968.4A5CD0BF@ina.fr> Tim Bumgarner wrote: > > What is the best way to prepare TIFF images to be used as icons, especially > in regards to the alpha channels? I've tried PhotoShop 5 and created an icon > with either a straight alpha channel or the quick mask feature and saved the > icon as a TIFF. When I use the icon in a MOXS app, I get the following > warning in the log window: > > Warning: TIFF image with unknown extra samples assumed to have unassociated > alpha. RGB values have been premultiplied > > The icon works fine, and has the correct Alpha (or mask). Anyone have any > ideas? I use GraphicConverter to build icon and it works fine, excepted that the way to define alpha is a bit tricky. Sincerely. -- G?rard Iglesias Ph.D. in Computer Graphic gerard.iglesias@ina.fr Researcher/Project Manager http://www.mediaport.net/People/Gerard.Iglesias From aep at apple.com Mon May 10 12:57:05 1999 From: aep at apple.com (Andrew Platzer) Date: Thu Nov 3 14:42:02 2005 Subject: TIFF images Message-ID: <199905101957.MAA25128@scv1.apple.com> > > Warning: TIFF image with unknown extra samples assumed to have unassociated > > alpha. RGB values have been premultiplied > > The icon works fine, and has the correct Alpha (or mask). Anyone have any > > ideas? Photoshop saves the alpha channel as an unknown channel (as far as the TIFF standard is concerned) and the code guesses that it is non-premultiplied alpha. The AppKit uses premultiplied (Porter-Duff) alpha and this warning tells you that it is mutliplying every single RGB value in the image by the alpha each time it loads. You can use the 'tiffutil' program to read the original image and write it in premutiplied form. Andrew ______________________________________________________________ Andrew Platzer | Application Frameworks | Apple Computer, Inc. (408) 974-2875 | aep@apple.com From ahoesch at smartsoft.de Mon May 10 14:17:35 1999 From: ahoesch at smartsoft.de (Andreas =?iso-8859-1?Q?H=F6schler?=) Date: Thu Nov 3 14:42:02 2005 Subject: DPS Errors on OPENSTEP 4.2 Mach Message-ID: <9905102117.AA15974@smartsoft.de> A non-text attachment was scrubbed... Name: not available Type: text/enriched Size: 1142 bytes Desc: not available Url : /mailman/archive/macosx-dev/attachments/19990510/8df7f045/attachment.bin From benster at wenet.net Mon May 10 15:41:12 1999 From: benster at wenet.net (James M. Benster) Date: Thu Nov 3 14:42:02 2005 Subject: Yellow Box BOF details available? Message-ID: <199905102239.PAA21147@mom.hooked.net> Are details available for the Yellow Box birds-of-a-feather get-together that Malcolm had mentioned was being scheduled for Mon. night? I just checked StepWise and saw nothing there. --Jim From bbum at codefab.com Mon May 10 16:22:45 1999 From: bbum at codefab.com (Bill Bumgarner) Date: Thu Nov 3 14:42:02 2005 Subject: Yellow Box BOF details available? In-Reply-To: <199905102239.PAA21147@mom.hooked.net> Message-ID: A little late now... but: The BOF is tonight starting at 6pm at the hilton near the convention center. Alternatively, meet in front of the convention center near the fountain at 5:30pm. On Mon, 10 May 1999, James M. Benster wrote: > Are details available for the Yellow Box birds-of-a-feather get-together > that Malcolm had mentioned was being scheduled for Mon. night? I just > checked StepWise and saw nothing there. > > --Jim > > From kc at omnigroup.com Mon May 10 16:24:17 1999 From: kc at omnigroup.com (Ken Case) Date: Thu Nov 3 14:42:02 2005 Subject: Yellow Box BOF details available? Message-ID: <199905102324.QAA04761@ignem.omnigroup.com> The Yellow Box BOF is at 6pm at the Hilton in Alameda room 2. From ken_anderson at mindspring.com Mon May 10 16:22:07 1999 From: ken_anderson at mindspring.com (Ken Anderson) Date: Thu Nov 3 14:42:02 2005 Subject: Yellow Box BOF details available? Message-ID: <01BE9B1C.3EB03F00@user-33qsqgj.dialup.mindspring.com> It will be held at 6 PM in Almaden 2 in the Hilton hotel. Ken -----Original Message----- From: James M. Benster [SMTP:benster@wenet.net] Sent: Monday, May 10, 1999 6:46 PM To: Multiple recipients of list Subject: Yellow Box BOF details available? Are details available for the Yellow Box birds-of-a-feather get-together that Malcolm had mentioned was being scheduled for Mon. night? I just checked StepWise and saw nothing there. --Jim From jman at clubi.ie Mon May 10 17:02:34 1999 From: jman at clubi.ie (justin armstrong) Date: Thu Nov 3 14:42:02 2005 Subject: cocoa? Message-ID: <3737739A.9E1A662E@clubi.ie> I can't believe they've renamed YB as Cocoa. YB was a bad enough name. Cocoa is just...embarassing. I'm going back to calling it OpenStep... From btoback at optc.com Mon May 10 17:27:07 1999 From: btoback at optc.com (Bruce Toback) Date: Thu Nov 3 14:42:02 2005 Subject: cocoa? Message-ID: <199905110032.RAA15839@landru.optc.com> Justin Armstrong writes: >I can't believe they've renamed YB as Cocoa. >YB was a bad enough name. Cocoa is Apple's graphical animation programming environment for kids. Steve must have gotten mixed up. Or maybe he meant Koko, as in the gorilla who gained fame in the '80s as the experimental subject in extensive research into primate language. Yeah, that's it. -- Bruce -------------------------------------------------------------------------- Bruce Toback Tel: (602) 996-8601| My candle burns at both ends; OPT, Inc. (800) 858-4507| It will not last the night; 11801 N. Tatum Blvd. Ste. 142 | But ah, my foes, and oh, my friends - Phoenix AZ 85028 | It gives a lovely light. btoback@optc.com | -- Edna St. Vincent Millay From nate at portents.com Mon May 10 17:39:59 1999 From: nate at portents.com (Nathan Raymond) Date: Thu Nov 3 14:42:02 2005 Subject: cocoa? In-Reply-To: <199905110032.RAA15839@landru.optc.com> References: <199905110032.RAA15839@landru.optc.com> Message-ID: At 5:33 PM -0700 5/10/99, Bruce Toback wrote: > >Justin Armstrong writes: > > >I can't believe they've renamed YB as Cocoa. > >YB was a bad enough name. > >Cocoa is Apple's graphical animation programming environment for kids. >Steve must have gotten mixed up. The guy who headed the Cocoa project left Apple to develop and market Cocoa seperately - that was almost a year ago, I think. So that Cocoa is no longer part of Apple. -- Nathan Raymond From eike at orang.org Mon May 10 18:26:36 1999 From: eike at orang.org (eike@orang.org) Date: Thu Nov 3 14:42:02 2005 Subject: porting faq: dynamic linking? (rld_load, dyld) Message-ID: <9905110326.AA366242@moon> I need to port some stuff that uses rld_load() in the NS3 port (sicstus prolog) It seems rld was replaced by dyld in macosx but I couldn't find much docs except /System/Library/Frameworks/System.framework/Headers/mach-o/dyld.h Can anyone point me to more docs/example code for the _dyld stuff? What is the proposed interface for dynamic linking in macosxs? ~eike ------ What is a computer? A stupid machine that only knows two states (power on/power off) -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: text/enriched Size: 533 bytes Desc: not available Url : /mailman/archive/macosx-dev/attachments/19990511/00a0c522/attachment.bin From eike at orang.org Mon May 10 18:26:34 1999 From: eike at orang.org (eike@orang.org) Date: Thu Nov 3 14:42:02 2005 Subject: porting faq: pthreads port? (was Porting mySQL to OS X Server (Help!)) Message-ID: <9905110326.AA341701@moon> It seems the main problem with MySQL is the missing support for pthreads in MacOSXServer. The freely pthread support doesn't help much since it is really bound to the underlying os thread implementations which are quite different from the Mach foundation of osx. If you try to port the pthreads package anyway, take a look into the darwin code. There's some code for implementing locks (with workarounds for various ppc bugs ...) Maybe some cthreads based wrapper would help but I don't have real specs for pthreads (the spec is not available for free ...) which makes things a bit vague. It seems that MySQL is only using a small subset of pthreads so maybe things can be done without a full port. If anyone is working on pthreads, please get in contact. ~eike ------ What is a computer? A stupid machine that only knows two states (power on/power off) -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: text/enriched Size: 938 bytes Desc: not available Url : /mailman/archive/macosx-dev/attachments/19990511/4804e743/attachment.bin From jray at postoffice.ag.ohio-state.edu Mon May 10 20:43:59 1999 From: jray at postoffice.ag.ohio-state.edu (John Ray) Date: Thu Nov 3 14:42:02 2005 Subject: Mac OS X questions... [namely, when do us non-attendees get the goods?] Message-ID: Has Apple indicated if/when those of us who aren't attending WWDC will see the preview release 1 of Mac OS X? I certainly hope it's included in the May or June seedings... For those who have it, any comments on the development environment? (I'm assuming it includes PB, etc.) Is carbon development supported within PB? Does it support NetInfo? Will I be able to integrate an OS X development machine into our OS X Server NetInfo group? Just anxious... Thanks, John Ray System Dev/Eng, OSUE Communications & Technology From malte at oops.se Tue May 11 03:54:13 1999 From: malte at oops.se (Malte Tancred) Date: Thu Nov 3 14:42:02 2005 Subject: Registering associations with IB Message-ID: <9905111254.AA130052@ostrawberry> Hi, I've created a custom association for a custom interface class and would like the association to appear in InterfaceBuilder. The GUI class is placed in a palette together with the association. How do I make IB aware of the assocation, ie make the assocation show up as an alternative in the association pop up in the connections inspector? Cheerio, Malte -- Malte Tancred, OOPS ab malte@oops.se http://www.oops.se From malte at oops.se Tue May 11 04:11:42 1999 From: malte at oops.se (Malte Tancred) Date: Thu Nov 3 14:42:02 2005 Subject: SOLVED: Re: Registering associations with IB In-Reply-To: <9905111254.AA130052@ostrawberry> Message-ID: <9905111311.AA428131@ostrawberry> I didn't override the method +displayName (which I didn't believe I had to). But now it works. Thanks! :-) Malte -- Malte Tancred, OOPS ab malte@oops.se http://www.oops.se -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: text/enriched Size: 621 bytes Desc: not available Url : /mailman/archive/macosx-dev/attachments/19990511/62409858/attachment.bin From tom at evatac.com Tue May 11 07:33:18 1999 From: tom at evatac.com (Tom Carstensen) Date: Thu Nov 3 14:42:02 2005 Subject: Mac OS X questions... [namely, when do us non-attendees get the Message-ID: <1285689287-21753034@evatac.com> >Has Apple indicated if/when those of us who aren't attending WWDC will see >the >preview release 1 of Mac OS X? I certainly hope it's included in the May or >June seedings... > >For those who have it, any comments on the development environment? (I'm >assuming it includes PB, etc.) Is carbon development supported within PB? > >Does it support NetInfo? Will I be able to integrate an OS X development >machine into our OS X Server NetInfo group? My question is why us Apple Developers not a WWDC don't have access to this. The dev ftp site has no sign of it. -Tom Carstensen From lavoie at cst.ca Tue May 11 08:45:48 1999 From: lavoie at cst.ca (Martin-Gilles) Date: Thu Nov 3 14:42:02 2005 Subject: Mac OS X questions... [namely, when do us non-attendees get the Message-ID: <199905111545.LAA16218@plexus.cst.ca> >My question is why us Apple Developers not a WWDC don't have access to >this. The dev ftp site has no sign of it. > >-Tom Carstensen That's an incentive for you to register for next year's WWDC. Apple has always given gifts of all sorts at the 'DC, and experimental/beta system software & components has always been part of the joy ride. The software stuff eventually cascades down to those developers who dont have the chance to go to the 'DC. Martin-Gilles Lavoie | "No! Try not. Do, or do not. Mac OS programmer, CS&T | There is no try." lavoie@cst.ca | -- Yoda on error handling http://blackhole.cst.ca/ | Y2K statement: hahahaha! From nhoj at cd.chalmers.se Tue May 11 09:31:03 1999 From: nhoj at cd.chalmers.se (nhoj@cd.chalmers.se) Date: Thu Nov 3 14:42:02 2005 Subject: non-rectangular clippaths for EPS images do not work on screen In-Reply-To: <199905100827.KAA18538@stsrv.select-tech.si> References: <199905091414.QAA26205@stsrv.select-tech.si> <37369654.B94BD0AE@ina.fr> <199905100827.KAA18538@stsrv.select-tech.si> Message-ID: <19990511163104.28925.qmail@glenn> You wrote: > If this image was created from EPS data, then on screen it is clipped to > bounding box of clip path, not to clip path !!! I guess this has something > to do with internal image cacheing/display mechanism (bug?) of > NSEPSImageRep and window server, because with TIFF images everything works > OK. In the GraphicKit, I clip to arbitrary NSBezierPaths with no problems when compositing NSImages containing EPS or tiff data to the screen. This works on RDR2, at least. Beta testers are welcome, by the way. First beta should be in a few days. OAAI's GlyphiX used to have the problem you're describing, by the way. Maury and I discussed it at length, and he ended up performing the clipping when doing the rendering, IIRC. Solves the problem, but I think he was a bit reluctant to go that way since it's kind of a kludge. Regards, John Hornkvist From smithja at cs.unc.edu Tue May 11 10:04:02 1999 From: smithja at cs.unc.edu (Jason M. Smith) Date: Thu Nov 3 14:42:02 2005 Subject: Mac OS X questions... [namely, when do us non-attendees get In-Reply-To: <199905111545.LAA16218@plexus.cst.ca> Message-ID: On Tue, 11 May 1999, Martin-Gilles wrote: > >My question is why us Apple Developers not a WWDC don't have access to > >this. The dev ftp site has no sign of it. > > > >-Tom Carstensen > > That's an incentive for you to register for next year's WWDC. > > Apple has always given gifts of all sorts at the 'DC, and experimental/beta > system software & components has always been part of the joy ride. > > The software stuff eventually cascades down to those developers who dont > have the chance to go to the 'DC. Unless you're a student developer, of course... no FTP access to the dev sites at all, a 'restricted' monthly mailing = it ain't monthly, and it's usually nothing more than updates that you could get from Apple's public download sites a month before, etc, etc. Ah well. $100 spent, lesson learned. Next year it'll be time to suck up and pay full ride. Jason Smith 333 Sitterson Hall smithja@cs.unc.edu (919) 962-1821 ------------------------------------------------------------------ http://www.cs.unc.edu/~smithja/ From ash at sineware.com Tue May 11 20:31:33 1999 From: ash at sineware.com (Mishra, Ashish) Date: Thu Nov 3 14:42:02 2005 Subject: creating boot-scripts Message-ID: <199905120230.WAA19246@engmail.uwaterloo.ca> Hello, Hopefully someone can help / point me in the right direction here... I'm trying to set up two custom Forth words in OpenFirmware 3.1.1: mac: to boot Mac OS ten: to boot OS X Server I know how to create the words, but not what the contents should be. Any examples would be appreciated. thanks in advance! Ash From norbert.heger at obdev.at Wed May 12 02:42:02 1999 From: norbert.heger at obdev.at (Norbert Heger) Date: Thu Nov 3 14:42:02 2005 Subject: cocoa? References: <3737739A.9E1A662E@clubi.ie> Message-ID: <9905120942.AA00399@cray> > I can't believe they've renamed YB as Cocoa. > YB was a bad enough name. Cocoa is just...embarassing. Why not rename the whole company? Seems consequent to me ;-) [An attachment was originally included here] Sorry for the bandwidth, but I couldn't resist... Norbert _____________________________ _________________________________ O B J E C T I V E D E V E L O P M E N T Norbert Heger Cocoa MacOS-X EOF WebObjects norbert.heger@obdev.at Software Development +43-664-2811777 www.obdev.at -------------- next part -------------- Skipped content of type multipart/mixed From lavoie at cst.ca Wed May 12 07:00:33 1999 From: lavoie at cst.ca (Martin-Gilles) Date: Thu Nov 3 14:42:02 2005 Subject: cocoa? Message-ID: <199905121400.KAA21396@plexus.cst.ca> > >Sorry for the bandwidth, but I couldn't resist... What's that got to do with anything? Martin-Gilles Lavoie | "No! Try not. Do, or do not. Mac OS programmer, CS&T | There is no try." lavoie@cst.ca | -- Yoda on error handling http://blackhole.cst.ca/ | Y2K statement: hahahaha! From penrose at sfc.keio.ac.jp Wed May 12 07:08:14 1999 From: penrose at sfc.keio.ac.jp (Christopher Penrose) Date: Thu Nov 3 14:42:02 2005 Subject: cocoa? In-Reply-To: <9905120942.AA00399@cray> Message-ID: At 5:08 AM -0700 5/12/99, Norbert Heger wrote: >> I can't believe they've renamed YB as Cocoa. >> YB was a bad enough name. Cocoa is just...embarassing. >Why not rename the whole company? >Seems consequent to me ;-) They should perhaps, to Sun Microsystems. Christopher penrose@sfc.keio.ac.jp From bbum at codefab.com Wed May 12 08:15:20 1999 From: bbum at codefab.com (Bill Bumgarner) Date: Thu Nov 3 14:42:02 2005 Subject: cocoa? In-Reply-To: <9905120942.AA00399@cray> Message-ID: Cocoa = Core Objective-C Object Architecture Courtesy of Ken Anderson... :-) From jean-michel.cazaux at finindev.com Wed May 12 08:19:11 1999 From: jean-michel.cazaux at finindev.com (Jean-Michel Cazaux) Date: Thu Nov 3 14:42:02 2005 Subject: Tabstop in a popup ? Message-ID: <01BE9C9B.8E5399A0@JEAN-MICHEL> Any idea on how to place a tabstop in a popup ? (the popup is filled with an EOassociation) Thanks in advance for any tip !!! __________________________________________________________________________________________ Jean-Michel CAZAUX FININDEV, Conseil en Finances pour les Collectivit?s Locales. 204 Rue Michel Teule - ZAC d'Alco 34080 Montpellier - FRANCE. T?l. +33 (0)4 67 63 66 25 - Fax +33 (0)4 67 63 35 45 e-mail: jean-michel.cazaux@finindev.com From eric at alum.mit.edu Wed May 12 08:39:07 1999 From: eric at alum.mit.edu (Eric Hermanson) Date: Thu Nov 3 14:42:02 2005 Subject: cocoa? References: Message-ID: <3739A09B.5B90EDA4@alum.mit.edu> Bill Bumgarner wrote: > > Cocoa = Core Objective-C Object Architecture > > Courtesy of Ken Anderson... > > :-) Wow! I hope that's not just a figurative statement. If they plan on keeping Obj-C, that would be great. Eric From paipai at tin.it Wed May 12 10:18:07 1999 From: paipai at tin.it (Paolo Di Francesco) Date: Thu Nov 3 14:42:02 2005 Subject: Darwin on Intel Message-ID: <19990512161329.NNHX3147.fep04-svc@harlock> Is someone working on it? Ciao Ciao Paolo Di Francesco _ ->B<- All Recycled Bytes Message ... ~ From ahoesch at smartsoft.de Wed May 12 11:00:04 1999 From: ahoesch at smartsoft.de (Andreas =?iso-8859-1?Q?H=F6schler?=) Date: Thu Nov 3 14:42:02 2005 Subject: Programmatically setting paragraphSpacing (OPENSTEP Text System) Message-ID: <9905121800.AA08477@smartsoft.de> A non-text attachment was scrubbed... Name: not available Type: text/enriched Size: 742 bytes Desc: not available Url : /mailman/archive/macosx-dev/attachments/19990512/15aed332/attachment.bin From smithja at cs.unc.edu Wed May 12 10:18:41 1999 From: smithja at cs.unc.edu (Jason M. Smith) Date: Thu Nov 3 14:42:02 2005 Subject: cocoa? In-Reply-To: Message-ID: At 8:31 AM -0700 5/12/99, Bill Bumgarner wrote: >Cocoa = Core Objective-C Object Architecture > >Courtesy of Ken Anderson... > >:-) Alright, that almost makes it better... but... Wouldn't the larger 'Yellow Box' environment then more properly be called Mocha? Java + Cocoa. :) I don't know about you, but Mocha sounds a little more... robust... than Cocoa. You program in the Mocha environment, and get your choice of Java, or Cocoa, or a mix of the two. Cocoa? *shudder* ------- -Jason From satadru at umich.edu Wed May 12 12:46:51 1999 From: satadru at umich.edu (Satadru Pramanik) Date: Thu Nov 3 14:42:02 2005 Subject: driverkit release Message-ID: I am hearing that the driver kit has finally been released for Darwin. if this is true, has anybody considered proting the Adaptec scsi/ultrascsi/ etc cards over to darwin/MOSXS? The drivers for Linux work Great, and it wouled be REALLY good (read essential) if MOSXS supported these cards. --= satadru pramanik `satadru@umich.edu' www.umich.edu/~satadru Online Editor, The Michigan Daily Yugoslavia declares 'peace' in Kosovo; NATO Airstrikes continue -CNN Interactive, 04/08/99 =-- From cpry at enc.org Wed May 12 12:45:51 1999 From: cpry at enc.org (chad pry) Date: Thu Nov 3 14:42:02 2005 Subject: security questions on macOS X Message-ID: <001801be9cb0$0a2d4a30$7b619280@enc.org> could someone point me to a tcp wrapper that can be installed under OS X. also, is there a compilable version of ssh out yet for OS X? chad From rasmussn at soren.lanl.gov Wed May 12 06:20:57 1999 From: rasmussn at soren.lanl.gov (Craig E Rasmussen) Date: Thu Nov 3 14:42:02 2005 Subject: security questions on macOS X In-Reply-To: <001801be9cb0$0a2d4a30$7b619280@enc.org> (cpry@enc.org) References: <001801be9cb0$0a2d4a30$7b619280@enc.org> Message-ID: <199905121320.HAA17489@soren.lanl.gov> >also, is there a compilable version of ssh out yet for OS X? I configured and built ssh-1.2.26 without a problem using ./configure --host=nextstep --disable-asm Craig From inphlux at lotek.org Wed May 12 13:31:58 1999 From: inphlux at lotek.org (ix) Date: Thu Nov 3 14:42:02 2005 Subject: security questions on macOS X Message-ID: <199905121949.PAA17176@project.lotek.org> > could someone point me to a tcp wrapper that can be installed under OS X. ftp://ftp.bow.net//pub/rhapsody/tcp_wrappers.tar.gz From eb at object.com Wed May 12 15:00:36 1999 From: eb at object.com (Eric Bergerson) Date: Thu Nov 3 14:42:02 2005 Subject: CDROM Burners for software distribution on NT, MacOSX, OpenStep Message-ID: <9905122200.AA00999@object.com> Can anyone suggest the best method, hardware, and software to use to press CDROMs for distributing YB software on MacOSX, NT and OpenStep. Sincerely, Eric Bergerson Objective Technologies, Inc. email: eb@object.com voice: 212 988 6268 fax: 212 988 6273 web: for those not on the talk list, the bomb finally dropped: http://www.stepwise.com/SpecialCoverage/WWDC99/index.html In response to a question about whether or not the new graphics architecture would be accessible from WebObjects under NT, Ken Bereskin let this out of the bag. -- At this time, we have no formal commitment to Cocoa on any platform other than Mac OS X Server _____________________________________ Glenn Carnagey Webmaster Chicago Sun-Times / Digital Chicago 401 N. Wabash, Rm. 513 Chicago, IL 60611 glenn@suntimes.com, www.suntimes.com 312.321.2603, fax 312.321.2849 _____________________________________ From dyoung at vviuh221.vvi.com Wed May 12 16:40:37 1999 From: dyoung at vviuh221.vvi.com (David Young) Date: Thu Nov 3 14:42:02 2005 Subject: death of YB/Windows References: <4.2.0.37.19990512175411.00b427e0@pop.interaccess.com> Message-ID: <9905122340.AA12369@vviuh221.vvi.com> > -- At this time, we have no formal commitment to Cocoa on any > platform other than Mac OS X Server Probably another Ken Bereskin foot in mouth. I suspect he meant or said "Mac OS X and Mac OS X Server", i.e. not just Server. Anyways... http://www.vvi.com/PressReleases/990505OGWIN32.html Thanks A Bunch! David Young; VVI-DCS dyoung@vvi.com From mtrent at msn.fullfeed.com Wed May 12 16:39:10 1999 From: mtrent at msn.fullfeed.com (mtrent@msn.fullfeed.com) Date: Thu Nov 3 14:42:02 2005 Subject: death of YB/Windows In-Reply-To: <4.2.0.37.19990512175411.00b427e0@pop.interaccess.com> from "Glenn Carnagey" at May 12, 99 04:04:56 pm Message-ID: <199905122339.SAA26076@shell1.msn.fullfeed.com> So, you interpret "At this time" to mean "Never at all"? Why not take Ken's comments at face value? My opinion. Mike Trent "In a previous message you (Glenn Carnagey) write:" > > > for those not on the talk list, the bomb finally dropped: > > http://www.stepwise.com/SpecialCoverage/WWDC99/index.html > > In response to a question about whether or not the new graphics > architecture would be accessible from WebObjects under NT, Ken Bereskin > let this out of the bag. > > -- At this time, we have no formal commitment to Cocoa on any > platform other than Mac OS X Server > > > > > > _____________________________________ > Glenn Carnagey > Webmaster > Chicago Sun-Times / Digital Chicago > 401 N. Wabash, Rm. 513 > Chicago, IL 60611 > glenn@suntimes.com, www.suntimes.com > 312.321.2603, fax 312.321.2849 > _____________________________________ > > > From eric at EMIEng.com Wed May 12 16:27:07 1999 From: eric at EMIEng.com (Eric Marshall) Date: Thu Nov 3 14:42:02 2005 Subject: looking for API to embed an application in the menu bar Message-ID: <199905122327.TAA26155@bobo.shirenet.com> I assume there's some new API to embed an application in the menu bar, ala the Clock app. I did some finds for "menubar" and "menu bar", but didn't find anything promising. Any pointers would be appreciated, or any explanation of how it's done without new API. Thanks in advance. From epeyton at epicware.com Wed May 12 17:13:08 1999 From: epeyton at epicware.com (Eric S. Peyton) Date: Thu Nov 3 14:42:02 2005 Subject: looking for API to embed an application in the menu bar In-Reply-To: <199905122327.TAA26155@bobo.shirenet.com> Message-ID: NSStatusBar and NSStatusMenu*. No docs but they are included in header form (very self explanatory) in AppKit. Eric Eric Peyton epeyton@epicware.com Software and Source for Mac OS X Server On Wed, 12 May 1999, Eric Marshall wrote: > I assume there's some new API to embed an application in the menu bar, > ala the Clock app. I did some finds for "menubar" and "menu bar", but didn't > find anything promising. Any pointers would be appreciated, or any explanation > of how it's done without new API. > > Thanks in advance. > > From eric at EMIEng.com Wed May 12 18:44:25 1999 From: eric at EMIEng.com (Eric Marshall) Date: Thu Nov 3 14:42:02 2005 Subject: looking for API to embed an application in the menu bar Message-ID: <199905130144.VAA28499@bobo.shirenet.com> > I assume there's some new API to embed an application in the menu bar, > ala the Clock app. I did some finds for "menubar" and "menu bar", but didn't > find anything promising. Any pointers would be appreciated, or any explanation > of how it's done without new API. Following up my own post, I found the source code for MenuClock app, from one of the DR eras, and it simply *positions* a borderless window *above* the menu bar. Not as sophisticated as I was assuming, but apparently it gets the job done :-) I'd still be interested in hearing about a preferred technique, if one exists. From bentley at crenelle.com Wed May 12 19:09:43 1999 From: bentley at crenelle.com (Michael Brian Bentley) Date: Thu Nov 3 14:42:02 2005 Subject: death of YB/Windows In-Reply-To: <199905122339.SAA26076@shell1.msn.fullfeed.com> Message-ID: >So, you interpret "At this time" to mean "Never at all"? Why not take >Ken's comments at face value? > >My opinion. > >Mike Trent Thoughtful comment. However, continuing this thread here is bad... -m From brianw at irisgroup.com Wed May 12 19:23:49 1999 From: brianw at irisgroup.com (Brian Wagner) Date: Thu Nov 3 14:42:02 2005 Subject: JPEG codex Message-ID: <373A37B4.10374E13@irisgroup.com> The DR2 notes stated there was now support for JPEG encoding. I can't find any info on it. Does anybody know about it. From epeyton at epicware.com Wed May 12 22:18:49 1999 From: epeyton at epicware.com (Eric S. Peyton) Date: Thu Nov 3 14:42:02 2005 Subject: looking for API to embed an application in the menu bar In-Reply-To: <199905130144.VAA28499@bobo.shirenet.com> Message-ID: I wrote MenmuClock and it wasn't the right way to do it. NSStatusItem is ... check it out. I know Caffeine is using it for a NSStatusItem version of their incredibly cool cpu load meter Cycles5.app (look on Caffeine's ftp site for a copy) (ftp.caffeine.com/pub/other - or something like that ...). Eric Eric Peyton epeyton@epicware.com Software and Source for Mac OS X Server On Wed, 12 May 1999, Eric Marshall wrote: > > I assume there's some new API to embed an application in the menu bar, > > ala the Clock app. I did some finds for "menubar" and "menu bar", but didn't > > find anything promising. Any pointers would be appreciated, or any explanation > > of how it's done without new API. > > Following up my own post, I found the source code for MenuClock app, from > one of the DR eras, and it simply *positions* a borderless window *above* the menu > bar. Not as sophisticated as I was assuming, but apparently it gets the job done :-) > I'd still be interested in hearing about a preferred technique, if one exists. > > From brianw at sounds.WA.COM Wed May 12 22:38:42 1999 From: brianw at sounds.WA.COM (Brian Willoughby) Date: Thu Nov 3 14:42:02 2005 Subject: security questions on macOS X In-Reply-To: <199905121949.PAA17176@project.lotek.org> References: <199905121949.PAA17176@project.lotek.org> Message-ID: <199905130538.WAA07018@sounds.WA.COM> [ > could someone point me to a tcp wrapper that can be installed [ > under OS X. [ [ ftp://ftp.bow.net//pub/rhapsody/tcp_wrappers.tar.gz What about the one in Darwin? I assume this means its part of Mac OS X Server (I haven't looked because I don't need tcp_wrappers at the moment), or at least compiles and works... http://www.publicsource.apple.com/projects/source/tcp_wrappers-7.6-1.tar.gz Brian Willoughby Software Design: NEXTSTEP, OpenStep, Rhapsody, MacOS X Sound Consulting Apple Enterprise Alliance Partner NeXTmail welcome From gerard.iglesias at ina.fr Thu May 13 02:15:24 1999 From: gerard.iglesias at ina.fr (Gerard Iglesias) Date: Thu Nov 3 14:42:02 2005 Subject: JPEG codex References: <373A37B4.10374E13@irisgroup.com> Message-ID: <373A982C.477B151A@ina.fr> Brian Wagner wrote: > > The DR2 notes stated there was now support for JPEG encoding. > I can't find any info on it. Does anybody know about it. I you are on NT, I have the solution for this problem using Quicktime/NT. Sincerely. Gerard From ddavidso at apple.com Thu May 13 08:33:55 1999 From: ddavidso at apple.com (Douglas Davidson) Date: Thu Nov 3 14:42:02 2005 Subject: security questions on macOS X Message-ID: <199905131533.IAA36822@scv2.apple.com> Mac OS X Server ships with a version of tcp_wrappers (/usr/libexec/tcpd). If you don't like our choice of compile-time options, the source is included in the source package, or you can get it from Darwin. Douglas Davidson From andrew at swcp.com Thu May 13 08:37:40 1999 From: andrew at swcp.com (Andrew Stone) Date: Thu Nov 3 14:42:02 2005 Subject: cocoa? Message-ID: <199905131537.JAA15568@llama.swcp.com> Cocoa = C, Objective C Or Anything From mike at lorax.com Thu May 13 08:48:40 1999 From: mike at lorax.com (Mike Ferris) Date: Thu Nov 3 14:42:02 2005 Subject: JPEG codex Message-ID: <199905131643.JAA22706@boom.lorax.com> > > The DR2 notes stated there was now support for JPEG encoding. > I can't find any info on it. Does anybody know about it. JPEG, GIF, PNG, and TIFF are all supported by NSImage automatically. Just create an image from a JPEG file the same way you would from a TIFF file. You can use API in NSBitmapImageRep for more advanced operations than simply reading an image from a file (for instance, if you want to extract the contents of a bitmap image rep as JPEG data). Mike Ferris From truong at netmagic.net Thu May 13 09:43:37 1999 From: truong at netmagic.net (Thai Truong) Date: Thu Nov 3 14:42:02 2005 Subject: JPEG codex In-Reply-To: <373A37B4.10374E13@irisgroup.com> Message-ID: Use 'TIFFRepresentationUsingCompression:factor:' from NSImage and specify JPEG compression. I don't remember the specifics but you can check the header files. >The DR2 notes stated there was now support for JPEG encoding. >I can't find any info on it. Does anybody know about it. -Thai Truong From ahorovit at apple.com Thu May 13 14:32:16 1999 From: ahorovit at apple.com (Alex Horovitz) Date: Thu Nov 3 14:42:02 2005 Subject: WebObjects NEEDS your help! Message-ID: <199905132127.OAA53262@scv4.apple.com> ***NOT AN OFFICIAL APPLE ANNOUNCEMENT*** This is the last week in the Java Developers Journal's Reader's Choice Award and this afternoon, sadly, Sybase Enterprise Application Server has pulled ahead of WebObjects. They have had a big push in votes over the last 24 hours. I need each of you to dig deep one more time! Let's get a big vote out by friends of WebObjects everywhere! Please encourage your friends, family, and co-workers to vote for WebObjects via this URL: http://www.sys-con.com/java/readerschoice/nominationform.html The polls close on May 15th so we need people to vote NOW! Thank you for all that you have done to help us capture this award for Apple Computer and WebObjects! Best Regards, Alex ---- Alex Horovitz A Regular Guy Who Just Loves WebObjects! http://www.apple.com/webobjects From kevinh at wamnet.com Thu May 13 14:46:58 1999 From: kevinh at wamnet.com (Kevin Hoogheem) Date: Thu Nov 3 14:42:02 2005 Subject: WebObjects NEEDS your help! References: <199905132127.OAA53262@scv4.apple.com> Message-ID: <373B4852.8BDB8CEC@wamnet.com> Alex Horovitz wrote: > > ***NOT AN OFFICIAL APPLE ANNOUNCEMENT*** > > This is the last week in the Java Developers Journal's Reader's Choice Award > and this afternoon, sadly, Sybase Enterprise Application Server has pulled > ahead of WebObjects. They have had a big push in votes over the last 24 hours. > > I need each of you to dig deep one more time! > > Let's get a big vote out by friends of WebObjects everywhere! Please encourage > your friends, family, and co-workers to vote for WebObjects via this URL: > > http://www.sys-con.com/java/readerschoice/nominationform.html > > The polls close on May 15th so we need people to vote NOW! > > Thank you for all that you have done to help us capture this award for Apple > Computer and WebObjects! > > Best Regards, > Boy you would think this guys gets a bonus if WO finishes first ;)- -- Kevin Hoogheem WAM!NET Inc. Systems Operations 6100 West 110th St. V: 612-886-5057 Bloomington, MN 55438 F: 612-885-0687 E: kevinh@wamnet.com From karmat at matthias.org Thu May 13 18:49:44 1999 From: karmat at matthias.org (Karl N. Matthias) Date: Thu Nov 3 14:42:02 2005 Subject: security questions on macOS X In-Reply-To: <001801be9cb0$0a2d4a30$7b619280@enc.org> Message-ID: I was also able to get ssh to compile using this line: --host=powerpc-unknown-openbsd2.4 Karl On Wed, 12 May 1999, chad pry wrote: > could someone point me to a tcp wrapper that can be installed under OS X. > > also, is there a compilable version of ssh out yet for OS X? > > chad > > From moy at parc.xerox.com Thu May 13 19:17:56 1999 From: moy at parc.xerox.com (Edward Moy) Date: Thu Nov 3 14:42:02 2005 Subject: security questions on macOS X In-Reply-To: <001801be9cb0$0a2d4a30$7b619280@enc.org> Message-ID: On Wed, 12 May 1999, chad pry wrote: > also, is there a compilable version of ssh out yet for OS X? In addition to (C-based) ssh, you might try the Java version from: http://www.mindbright.se/mindterm/ To be honest, I haven't tried it on Mac OS X server, but I did get it to run on Mac OS 8.5.1 with MRJ 2.1.1 and JBindery (to make a double-clickable application). If it works with MRJ, I would expect it to work under OS X server as well. ------------------------------------------------------------------------ Edward Moy Xerox Palo Alto Research Center From lxs at MIT.EDU Thu May 13 21:14:23 1999 From: lxs at MIT.EDU (Alexandra Ellwood) Date: Thu Nov 3 14:42:02 2005 Subject: security questions on macOS X In-Reply-To: Your message of "Thu, 13 May 1999 19:31:48 PDT." Message-ID: <199905140414.AAA24932@ethanol> > On Wed, 12 May 1999, chad pry wrote: > > > also, is there a compilable version of ssh out yet for OS X? > > In addition to (C-based) ssh, you might try the Java version from: > > http://www.mindbright.se/mindterm/ > > To be honest, I haven't tried it on Mac OS X server, but I did get it to > run on Mac OS 8.5.1 with MRJ 2.1.1 and JBindery (to make a > double-clickable application). If it works with MRJ, I would expect it to > work under OS X server as well. I've compiled both ssh 1.2.26 and 2.0.12 for Mac OS X server. Unfortunately, being in the US, I can't submit my patches back to the ssh folks. However, the patches aren't substantial. Basically you just look for all the places they #ifdef NeXT and make sure that those are not defined for Mac OS X (like "#if (defined(NeXT) && !defined(__APPLE__))"). 1.2.x is harder if you want the assembly optimization... the mach assembler has different syntax from the powerpc32 assembly they have in the gmp library. But it's not much harder... just #define the registers and surround all the # with /* and you can get the assembly building too. Oh and you need to add Mac OS X to the config.guess and config.sub files... just steal some from some autoconfed Darwin thing. That should get it working just fine. Even the kerberos stuff works! I couldn't do my y2k testing securely without it! ;-) --lxs ----------------------------------------------------------------------------- Alexandra Ellwood MIT Information Systems http://web.mit.edu/lxs/www/index.html "Everybody dies frustrated and sad and that is beautiful." ----------------------------------------------------------------------------- From micha at object-factory.com Fri May 14 04:02:45 1999 From: micha at object-factory.com (Michael Maier) Date: Thu Nov 3 14:42:02 2005 Subject: AppleScript and YellowBox Message-ID: <199905141102.NAA00857@helicon.object-factory.com> hi all, I have got the following problem: I use AppleScript from the blue box to communicate to "TextEdit" like that: tell app "TextEdit" ... end tell that works fine. I save it as compiled script. After I logout, login and start TextEdit and MacOS.app and - of course - my compiled script, I get an error (error code -2). So I have to compile the script again und then it works...but I don't want to compile the script every time I launched MacOS.app...Has anybody an idea or knows a workaraound for that problem? thanks for any hints und tips Michael -- =========================================================== Michael Maier | | Object Factory GmbH | phone (0231)975137-0 Otto-Hahn-Strasse 18 | fax (0231)975137-99 44227 Dortmund | email m.maier@object-factory.com =========================================================== The software says Windows 98 or better, so I bought a Mac. From lavoie at cst.ca Fri May 14 06:19:04 1999 From: lavoie at cst.ca (Martin-Gilles) Date: Thu Nov 3 14:42:02 2005 Subject: WebObjects NEEDS your help! Message-ID: <199905141319.JAA08985@plexus.cst.ca> >***NOT AN OFFICIAL APPLE ANNOUNCEMENT*** > >This is the last week in the Java Developers Journal's Reader's Choice Award >and this afternoon, sadly, Sybase Enterprise Application Server has pulled >ahead of WebObjects. They have had a big push in votes over the last 24 hours. > >I need each of you to dig deep one more time! > >Let's get a big vote out by friends of WebObjects everywhere! Please encourage >your friends, family, and co-workers to vote for WebObjects via this URL: > >http://www.sys-con.com/java/readerschoice/nominationform.html > >The polls close on May 15th so we need people to vote NOW! > >Thank you for all that you have done to help us capture this award for Apple >Computer and WebObjects! > >Best Regards, > >Alex All right. Here's the deal: -We all jump in and make WO win But in return -Apple makes reasonable licensing fees for YB/Windows How's that? =========================================================================== Martin-Gilles Lavoie, Mac OS programmer, Corporate Software & Technologies --------------------------------------------------------------------------- (internet) http://pages.infinit.net/mouser/ mouser@videotron.ca (intranet) http://blackhole.cst.ca/ lavoie@cst.ca --------------------------------------------------------------------------- "No! Try not. Do, or do not. There is no try." -- Yoda on error handling From dyoung at vviuh221.vvi.com Fri May 14 06:45:45 1999 From: dyoung at vviuh221.vvi.com (David Young) Date: Thu Nov 3 14:42:02 2005 Subject: WebObjects NEEDS your help! References: <199905141319.JAA08985@plexus.cst.ca> Message-ID: <9905141345.AA14691@vviuh221.vvi.com> >All right. Here's the deal: > > -We all jump in and make WO win > >But in return > > -Apple makes reasonable licensing fees for YB/Windows > >How's that? I think that is a very thoughtful and realistic proposition, demonstrating a good working relationship between third party developers and Apple. Alex, what do you think? From markm at mail.tyrell.com Fri May 14 07:23:44 1999 From: markm at mail.tyrell.com (Mark F. Murphy) Date: Thu Nov 3 14:42:02 2005 Subject: Mac OS X questions... [namely, when do us non-attendees get In-Reply-To: <199905111545.LAA16218@plexus.cst.ca> Message-ID: At 8:57 AM -0700 5/11/99, Martin-Gilles wrote: >>My question is why us Apple Developers not a WWDC don't have access to >>this. The dev ftp site has no sign of it. >> >>-Tom Carstensen > >That's an incentive for you to register for next year's WWDC. > >Apple has always given gifts of all sorts at the 'DC, and experimental/beta >system software & components has always been part of the joy ride. > >The software stuff eventually cascades down to those developers who dont >have the chance to go to the 'DC. Well gee.... I'll just drop a week out of my busy development schedule on my current million dollar web site... Apple, just because I'm busy developing in the real world doesn't mean I'm any less worthy of getting timely developer releases that WWDC attendees get. Is there a reason I pay $500/yr for my membership? mark --------------------------------------------------------------------------- Mark F. Murphy, Director Software Development Tyrell Software Corp PowerPerl(tm), Add Power To Your Webpage! --------------------------------------------------------------------------- Member of Families Against Internet Censorship: rainbow.rmi.net/~fagin/faic From thomas.roehl at cai.com Fri May 14 07:35:38 1999 From: thomas.roehl at cai.com (Roehl, Thomas) Date: Thu Nov 3 14:42:02 2005 Subject: Mac OS X questions... [namely, when do us non-attendees get Message-ID: <4C094419B0E6D2119FF100805F85CABE70B06F@usilmse1.cai.com> I agree whole-heartedly. Has anyone heard any kind of news about whether the non-attendee developers will be getting the preview? And who else is upset that they are charging us for X Server? I really feel like my membership has been a complete waste of money... > ---------- > From: Mark F. Murphy[SMTP:markm@mail.tyrell.com] > Reply To: markm@mail.tyrell.com > Sent: Friday, May 14, 1999 10:29 AM > To: Multiple recipients of list > Subject: Re: Mac OS X questions... [namely, when do us non-attendees > get > > Well gee.... I'll just drop a week out of my busy development schedule on > my current million dollar web site... > > Apple, just because I'm busy developing in the real world doesn't mean I'm > any less worthy of getting timely developer releases that WWDC attendees > get. > > Is there a reason I pay $500/yr for my membership? > > mark > From jpgarner at erols.com Fri May 14 09:32:58 1999 From: jpgarner at erols.com (John P. Garner) Date: Thu Nov 3 14:42:02 2005 Subject: Mac OS X questions... [namely, when do us non-attendees get References: <4C094419B0E6D2119FF100805F85CABE70B06F@usilmse1.cai.com> Message-ID: <373C5037.7962@erols.com> Roehl, Thomas wrote: > > I agree whole-heartedly. Has anyone heard any kind of news about > whether the non-attendee developers will be getting the preview? > And who else is upset that they are charging us for X Server? I > really feel like my membership has been a complete waste of > money... > > > ---------- > > From: Mark F. Murphy[SMTP:markm@mail.tyrell.com] > > Reply To: markm@mail.tyrell.com > > Sent: Friday, May 14, 1999 10:29 AM > > To: Multiple recipients of list > > Subject: Re: Mac OS X questions... [namely, when do us non-attendees > > get > > > > Well gee.... I'll just drop a week out of my busy development schedule on > > my current million dollar web site... > > > > Apple, just because I'm busy developing in the real world doesn't mean I'm > > any less worthy of getting timely developer releases that WWDC attendees > > get. > > > > Is there a reason I pay $500/yr for my membership? > > > > mark > > This is the response I received from the Apple Developer Connection. "Hello John, Thank you for contacting Apple Developer Connection. I am sorry that I have no concrete information for you regarding Mac OS X developer release that was distributed during the WWDC99. Please stay tuned to your ADC Newsletter and your Seeding Update email for any information regarding the released software." The only reason I joined the ADC was to get early access to the Rhapsody\Mac OS X Server\Mac OS X software for my development projects. Like some other developers, I'm quite disappointed with the delays I've experienced in receiving the software. Access to Veronica and Sonata have been excellent, but my critical need is for Mac OS X development. -- John P. Garner GARNER & Associates 9418 F Spring House Lane Laurel, Maryland 20708 (301)-725-2384 jpgarner@erols.com From btoback at optc.com Fri May 14 09:41:58 1999 From: btoback at optc.com (Bruce Toback) Date: Thu Nov 3 14:42:02 2005 Subject: Mac OS X questions... [namely, when do us non-attendees get Message-ID: <199905141646.JAA22677@landru.optc.com> Thomas Roehl writes: >I agree whole-heartedly. Has anyone heard any kind of news about >whether the non-attendee developers will be getting the preview? I haven't heard, but in past years developer program members have received all (or almost all) the WWDC "goodies" within a month of the conference, usually in the next month's developer mailing. (The one exception was two years ago, when MacOS developers didn't get the copy of WebObjects that was distributed at WWDC.) >And who else is upset that they are charging us for X Server? Not I. They pay royalties on some of the software being shipped on that CD. (I may be biased since I've paid for an EAP membership, so I *did* get the CD for "free".) OSX Developer Preview 1 doesn't have the royalty problem, so you'll get it for free. >I >really feel like my membership has been a complete waste of >money... Assuming you're signed up at the $500 level, what you get is: 1. Access to developer seeds, 2. The monthly mailing, which gets you all MacOS system software, and saves you the time and trouble of finding and downloading piles of stuff from the FTP site. 3. Access to the compatibility testing labs, and 4. Two tech support calls. If you don't have a need for these, then yes, you wasted your money. -- Bruce -------------------------------------------------------------------------- Bruce Toback Tel: (602) 996-8601| My candle burns at both ends; OPT, Inc. (800) 858-4507| It will not last the night; 11801 N. Tatum Blvd. Ste. 142 | But ah, my foes, and oh, my friends - Phoenix AZ 85028 | It gives a lovely light. btoback@optc.com | -- Edna St. Vincent Millay From karmat at matthias.org Fri May 14 10:29:41 1999 From: karmat at matthias.org (Karl N. Matthias) Date: Thu Nov 3 14:42:02 2005 Subject: security questions on macOS X In-Reply-To: <199905140414.AAA24932@ethanol> Message-ID: Hmmm. I didn't have to go through any of that. I just did ./configure --host=powerpc-unknown-openbsd2.4 and it just compiled. Not a single problem. Karl On Thu, 13 May 1999, Alexandra Ellwood wrote: > > > On Wed, 12 May 1999, chad pry wrote: > > > > > also, is there a compilable version of ssh out yet for OS X? > > > > In addition to (C-based) ssh, you might try the Java version from: > > > > http://www.mindbright.se/mindterm/ > > > > To be honest, I haven't tried it on Mac OS X server, but I did get it to > > run on Mac OS 8.5.1 with MRJ 2.1.1 and JBindery (to make a > > double-clickable application). If it works with MRJ, I would expect it to > > work under OS X server as well. > > I've compiled both ssh 1.2.26 and 2.0.12 for Mac OS X server. > Unfortunately, being in the US, I can't submit my patches back to the > ssh folks. However, the patches aren't substantial. Basically you > just look for all the places they #ifdef NeXT and make sure that those > are not defined for Mac OS X (like "#if (defined(NeXT) && > !defined(__APPLE__))"). > > 1.2.x is harder if you want the assembly optimization... the mach > assembler has different syntax from the powerpc32 assembly they have > in the gmp library. But it's not much harder... just #define the > registers and surround all the # with /* and you can get the assembly > building too. > > Oh and you need to add Mac OS X to the config.guess and config.sub > files... just steal some from some autoconfed Darwin thing. > > > That should get it working just fine. Even the kerberos stuff works! > I couldn't do my y2k testing securely without it! ;-) > > --lxs > ----------------------------------------------------------------------------- > Alexandra Ellwood > MIT Information Systems http://web.mit.edu/lxs/www/index.html > "Everybody dies frustrated and sad and that is beautiful." > ----------------------------------------------------------------------------- > > From cmh at greendragon.com Fri May 14 10:49:03 1999 From: cmh at greendragon.com (Chris Hanson) Date: Thu Nov 3 14:42:02 2005 Subject: Mac OS X questions... [namely, when do us non-attendees get In-Reply-To: <199905141646.JAA22677@landru.optc.com> Message-ID: At 12:11 PM -0500 5/14/99, Bruce Toback wrote: > 4. Two tech support calls. Also, the right to purchase more as needed. (And in case there are non-developers reading, these are developer code-level support calls, not simply "my Mac doesn't work" type calls. :) Additional items: 5. A $100 coupon for Metrowerks. Pure ProjectBuilder developers may not care, but those of us who love and use CodeWarrior for MacOS development really appreciate this. 6. Apple has some sort of new developer leasing program for Select and Premiere program members. The terms aren't listed on the web site and I haven't called yet, so I don't know how good it is. 7. A cheaper price on Mac OS X Server ($100, which probably covers Apple's costs and shipping). This doesn't mean that we won't all be very disappointed with Apple if they try to charge developers for the release version of Mac OS X even though they won't have to pay royalties to anyone... From lxs at MIT.EDU Fri May 14 10:55:23 1999 From: lxs at MIT.EDU (Alexandra Ellwood) Date: Thu Nov 3 14:42:02 2005 Subject: security questions on macOS X Message-ID: <199905141755.NAA00498@ethanol> > Hmmm. I didn't have to go through any of that. I just did > > ./configure --host=powerpc-unknown-openbsd2.4 > > and it just compiled. Not a single problem. Ah, I'm confused. What I was referring to were my patches to DR2 to build ssh. I didn't realize that "NeXT" had been removed from the specs file for gcc in Mac OS X Server, but lo it has (yay!). So yeah, if you mod config.guess and config.sub it should just work (or brute force it to think it's some 4.4 BSD os). However, it's still the case that if you're trying to compile 1.2.x and you want the assembly optimizations, then you'll have to do some assembly hacking. I know that the gmp assembly tests still fail without modifications. -- lxs ----------------------------------------------------------------------------- Alexandra Ellwood MIT Information Systems http://web.mit.edu/lxs/www/index.html "Everybody dies frustrated and sad and that is beautiful." ----------------------------------------------------------------------------- From paulrs at lgs-systems.com Fri May 14 14:12:24 1999 From: paulrs at lgs-systems.com (Paul R. Summermatter) Date: Thu Nov 3 14:42:02 2005 Subject: Debugging Nightmares on OSXServer Message-ID: <9905142112.AA11784@slab> Guys, Is anyone having success debugging on OSXServer? I feel like I'm back on my old Trash80, because all of my debugging has been reduced to NSLogs. Specifically, our app consists of about 8 frameworks and 4 bundles. I have compiled some but not all of the frameworks as 'debug' (compiling all as debug seems to make the size of the image just too big to manage). Regardless, the rest of the frameworks and bundles are compiled as just frameworks and bundles without any stripping of symbols or other special compiler flags. My milage varies from debugging session to debugging session, but I have found the following problems: 1) I have no way to walk up the stack and access variables which have been passed in as arguments to the methods on the stack. The offsets which work on OS 4.2 and YB/NT ($fp+8 etc), don't seem to work on OSX. Are there any different offsets which will work? If not, how does one get arguments to methods which are on the stack especially when the methods are defined in Apple's frameworks? 2) When I am in my own code, even having built debug, I frequently cannot access variables which are defined within the scope of the method where I have broken. For example, take the following code: - (void)foo { NSString *myString = [NSString stringWithString:@"bar"]; if (/*something is true*/) { <---------- Break in gdb here ... If I try to 'po myString' I commonly get back . This always seems to happen whenever trying to access a singleton object as in 'po [LGSPrivilegesManager defaultManager]'. Is this common? Is there a workaround? Does anyone have any suggestions on how to improve this situation? Has everyone else been reduced to println debugging as well? Any suggestions or sympathies :-) are welcome! Regards, Paul --- Paul Summermatter LGS Systems, Inc. Medical Computing Division 15 TJ Gamester Ave Portsmouth, NH 03801-5871 (603) 433-9822 voice (603) 433-9818 fax (888) 898-6321 pager 8986321@skytel.com paging email paulrs@lgs-systems.com (NeXT or MIME Mail Welcome) http://www.lgs-systems.com From jjfeiler at relief.com Fri May 14 15:33:03 1999 From: jjfeiler at relief.com (John Jay Feiler) Date: Thu Nov 3 14:42:02 2005 Subject: Debugging Nightmares on OSXServer Message-ID: <199905140233.TAA19053@haque.relief.com> > 1) I have no way to walk up the stack and access variables which have > been passed in as arguments to the methods on the stack. The offsets which > work on OS 4.2 and YB/NT ($fp+8 etc), don't seem to work on OSX. Are there > any different offsets which will work? If not, how does one get arguments to > methods which are on the stack especially when the methods are defined in > Apple's frameworks? PPC doesn't pass arguments on the stack unless it absolutely has to. The first 6 integer-like arguments are passed in registers $r3 through $r8, and the first 6 double/float arguments are passed in $f3 through $f8. in an ObjC message: po ((id*)($r3)) prints the receiver p ((char *)($r4)) prints the selector po ((id*)($r5)) prints the first argument that is integer like as an object p (char *)(*((*($r3))+8)) prints the classname of the receiver Our friendly 'trace' call is this printf "[%s %s]\n", ( $r3 ? (char *)(*((*($r3))+8)) : "nil" ), ((char *)($r4)) > 2) When I am in my own code, even having built debug, I frequently > cannot access variables which are defined within the scope of the method > where I have broken. For example, take the following code: > > - (void)foo > { > NSString *myString = [NSString stringWithString:@"bar"]; > > if (/*something is true*/) { <---------- Break in gdb here > ... > > If I try to 'po myString' I commonly get back . This always > seems to happen whenever trying to access a singleton object as in 'po > [LGSPrivilegesManager defaultManager]'. Is this common? Yes > Is there a workaround? The compiler on MOSXS seems to be doing a lot of instruction reordering. I'm not sure if the compiler is just smarter about methods that are effectively 'const' (i.e. that have no side effects and can be safely reordered), or perhaps I'm dreaming. I usually do 'info local' to see what things look like. disassembling the method is often useful as well, once you get used to the ppc assembly opcodes. -- John Feiler jjfeiler@relief.com From krw at nobugz.com Fri May 14 15:43:49 1999 From: krw at nobugz.com (Randy Wigginton) Date: Thu Nov 3 14:42:02 2005 Subject: Mac OS 8.6 In-Reply-To: <199905141755.NAA00498@ethanol> Message-ID: I know Apple recommends against it, but I already installed 8.6 in my blue box partition. It seems fine to me -- anybody else out there try it yet? --randy From jjfeiler at relief.com Fri May 14 15:57:10 1999 From: jjfeiler at relief.com (John Jay Feiler) Date: Thu Nov 3 14:42:02 2005 Subject: Mac OS 8.6 Message-ID: <199905140257.TAA19526@haque.relief.com> > From: Randy Wigginton > I know Apple recommends against it, but I already installed 8.6 in my blue > box partition. It seems fine to me -- anybody else out there try it yet? > --randy I installed it on two machines before the TIL went up, and I haven't seen any problems yet. J -- John Feiler jjfeiler@relief.com From rcfa at cubiculum.com Fri May 14 15:55:23 1999 From: rcfa at cubiculum.com (Ronald C.F. Antony) Date: Thu Nov 3 14:42:02 2005 Subject: Debugging Nightmares on OSXServer In-Reply-To: <199905140233.TAA19053@haque.relief.com> References: <199905140233.TAA19053@haque.relief.com> Message-ID: <199905142255.SAA25794@kannix.cubiculum.com> > PPC doesn't pass arguments on the stack unless it absolutely has to. [...] > The compiler on MOSXS seems to be doing a lot of instruction > reordering. The joys of debugging on a RISC platform... Have you tried to remove the -O option from compilation? I notice that under OS-M and NS, the compiler always has the -O option, even if the -g is also present. If that doesn't work, I wish Apple would modify the compiler to allow a debugging mode in which all code is generated in as dumb-as-it-gets CISC fashion: one instruction at a time, everything on the stack, etc. Ronald ============================================================================== "The reasonable man adapts himself to the world; the unreasonable one persists in trying to adapt the world to himself. Therefore all progress depends on the unreasonable man." G.B. Shaw | rcfa@cubiculum.com | NeXT-mail welcome From driedel at home.com Fri May 14 16:18:22 1999 From: driedel at home.com (David P. Riedel) Date: Thu Nov 3 14:42:02 2005 Subject: Mac OS 8.6 References: <199905140257.TAA19526@haque.relief.com> Message-ID: <373CAF3D.5B0A76EA@home.com> I've been running MacOS 8.6 on my bluebox since b3 of 8.6 -- no problems. i think the TIL is referring to the disk image that osx sets up, not an external operating system on a real disk. dave riedel John Jay Feiler wrote: > > From: Randy Wigginton > > I know Apple recommends against it, but I already installed 8.6 in > my blue > > box partition. It seems fine to me -- anybody else out there try > it yet? > > --randy > > I installed it on two machines before the TIL went up, and I haven't > seen any problems yet. > > J > > -- > John Feiler > jjfeiler@relief.com From tom at evatac.com Fri May 14 16:46:21 1999 From: tom at evatac.com (Tom Carstensen) Date: Thu Nov 3 14:42:02 2005 Subject: Mac OS X questions... [namely, when do us non-attendees get Message-ID: <1285396912-5321042@evatac.com> > >I haven't heard, but in past years developer program members have >received all (or almost all) the WWDC "goodies" within a month of the >conference, usually in the next month's developer mailing. (The one >exception was two years ago, when MacOS developers didn't get the copy of >WebObjects that was distributed at WWDC.) I'm personally upset that WWDC attendees got it and we non-attendees did not. A month??? uh.. putting up on the developer ftp will take all of a few minutes. Personally i can do totally with the monthly cd mailing -- i get it when it's announced via ftp. There is *clearly* some discrimination here, has it's just as easy to ftp it as it is to hand it out at the conference. I agree -- what do we pay $500/yr for? We are less of developers becuase apple didn't get our $1000 for the WWDC? -Tom Carstensen -- Tom Carstensen tom@evatac.com President http://www.evatac.com Evatac Enterprises 877.287.8618 (toll free) Macintosh, Windows & Be Software Development From ash at sineware.com Fri May 14 19:03:29 1999 From: ash at sineware.com (Mishra, Ashish) Date: Thu Nov 3 14:42:02 2005 Subject: Dual booting with Open Firmware on B&W G3s Message-ID: <199905150102.VAA31153@engmail.uwaterloo.ca> For those of you who want to dual-boot into Mac OS or OS X Server on the new B&W G3s, without pressing the 'option' key, nor waiting for Mac OS to half boot up... here's a solution. I figured out how to write boot scripts in OF, then discovered they weren't saved in the NVRAM - meaning I lost the scripts on every reboot/restart. Doh! But there is a workaround! and here it is... My setup is a B&W G3 (400, 6 GB Ultra-dma drive) with OpenFirmware 3.1.1 I partitioned the internal drive into Mac OS (partition 5) and a OS X Server (partition 6). Using Open firmware (option-command O-F), you can specify which drive to boot of off. For example, to boot Mac OS (partition 5 on ata drive 1), type: 0 > boot /pci@80000000/pci-bridge@d/pci-ata@1/ata-4@0/@0:5,\\:tbxi And for OS X (partition 6 on ata drive 1), type: 0 > boot /pci@80000000/pci-bridge@d/pci-ata@1/ata-4@0/@0:6,\\:tbxi This works, but it is pretty hard to remember, (isn't it? :) There is a better solution. Use the defined aliases: Type devalias for the list of your devices. You should be able to identify the device your hard drive is connected to. My drive is stock, connected directly to ide1. the alias 'hd' reads: /pci/@d/pci-ata@1/ata-4@0/disk@0 this is the same as /pci@80000000/pci-bridge@d/pci-ata@1/ata-4@0 (how do you know? type 0 > dev hd pwd) Anyway, you can boot using the devalias 'hd': for example, to boot Mac OS (partition 5), type at the OF prompt: 0 > boot hd:5,\\:tbxi and for OS X Server (partition 6), type: 0 > boot hd:6,\\:tbxi Much simpler! The nice thing is you can set either Mac OS or OS X Server as the default using the System Disk application (under Mac OS), and then just type 'bye' to go into the default partition. (You can identify your default partition by typing 'printenv' and looking at the 'boot-device' variable). The only problem is that no matter what, the screen resolution is lost when going into either Mac OS native or OS X natively. If anyone knows how to keep these settings, please do pass it on :) regards, Ash > Hopefully someone can help / point me in the right direction here... > > I'm trying to set up two custom Forth words in OpenFirmware 3.1.1: > > mac: to boot Mac OS > ten: to boot OS X Server > > I know how to create the words, but not what the contents should be. > > Any examples would be appreciated. > > thanks in advance! > Ash ~~ Ashish Mishra, Sine of the Times, http://www.sineware.com ~~ mac2400, http://www.sineware.com/mac2400 The person who says it cannot be done should not interrupt the person doing it. - Chinese proberb From ckane at apple.com Fri May 14 18:57:57 1999 From: ckane at apple.com (Chris Kane) Date: Thu Nov 3 14:42:02 2005 Subject: Debugging Nightmares on OSXServer Message-ID: <007a01be9e76$5a90c840$9b0eca11@perspicuity.apple.com> jjfeiler@relief.com writes: >PPC doesn't pass arguments on the stack unless it absolutely has to. > The first 6 integer-like arguments are passed in registers $r3 >through $r8, and the first 6 double/float arguments are passed in $f3 >through $f8. in an ObjC message: Actually it's $r3 - $r10 and $f1 - $f13. And if there are mixed integer/pointer and float arguments, the next 2 general registers (r3 - r10) are consumed as well for placeholders of the float in the GPRs if there are some left. But the rest was correct. Structures as arguments have additional rules, and on the whole the rules are very complex and not worth knowing. The practical upshot is that arguments can only be viewed reliably at the point a method or function is entered. In optimized code, the compiler does liveness analysis--like it should--and if an argument (can be common with _cmd) is not needed any longer the compiler will use that register for something else. paulrs@lgs-systems.com wrote: >> 2) When I am in my own code, even having built debug, I frequently >> cannot access variables which are defined within the scope of the >> method Well, again liveness analysis will reuse places when variables used to be stored, so a "po" may actually attempt to po something other than the object you wanted, as gdb thinks the variable is stored one place, but that's already being used for something else. Also some other types of optimizations seem to shuffle things around without leaving a symbol information trail for gdb to use. Declare the local variable "volatile" to fix it somewhere and keep it around for the life of the method. Even this may not help with a local variable and really clever compiler. Of course, that will defeat optimization, so you don't want to leave that around generally. Turning off optimization in the build is the best bet. Set OPTIMIZE_BUILD_CFLAGS= (no value) in your Makefile.postamble or on the make command-line or in the PB options panel I think will work as well. Setting OPTIMIZE=NO in your Makefile.preamble might work as well, but that may have some additional side effects on the build, I'm not sure. Turning off optimization will generate very dumb code, which can be almost illegible to a human, with saves to the stack and whatnot. Optimization tends to produce much easier-to-read code, because it is more like a human would have just written. My free advice, Chris Kane Apple Computer, Inc. From hunter at lastonepicked.com Fri May 14 21:27:52 1999 From: hunter at lastonepicked.com (Hunter Hillegas) Date: Thu Nov 3 14:42:02 2005 Subject: Justifying the Apple Developer Connection Membership Message-ID: <19990515042752622.AAA660@nts1.webhostingprovider.com@[207.71.250.180]> Hello. I have been a Select member of ADC for one year. I subscribed to develop for Yellow Box. In that time, I received DR2 of Rhapsody plus the "privilege" to buy MOSXS at a discount. That's it. $500 for that... For most companies, this is no big deal. They have the money to spend. I am a student. I love the Yellow Box. I want to finish my application. My membership ends 5/31. I was told by Apple that to get MOSX PR1 I would have to re-subscribe. This is bullshit. I have gotten next to nothing from my membership fee. The student developer program doesn't include software seeds, so it is worthless. Anyone else in this boat? From jray at postoffice.ag.ohio-state.edu Fri May 14 22:43:33 1999 From: jray at postoffice.ag.ohio-state.edu (John Ray) Date: Thu Nov 3 14:42:02 2005 Subject: Mac OS X questions... [namely, when do us non-attendees get Message-ID: >>I haven't heard, but in past years developer program members have >>received all (or almost all) the WWDC "goodies" within a month of the >>conference, usually in the next month's developer mailing. (The one >>exception was two years ago, when MacOS developers didn't get the copy of >>WebObjects that was distributed at WWDC.) > > I'm personally upset that WWDC attendees got it and we non-attendees > did not. Well, I'm almost sorry I started this thread. Personally, I'm happy w/ my $500 membership. Obviously the membership has been centered around MacOS developers and will continue to be until Mac OS X replaces the current system. My question was simply when we would get it? Those of us who didn't attend 2 (or 3?) years ago had to contact Apple to get OpenStep 4.2 (Prelude to Rhapsody). I have no problem waiting a month, and I have no problem w/ WWDC attendees getting the software first... I just want to make sure that I make the necessary arrangements to get it sometime in the reasonably near future. --- John From jjfeiler at relief.com Fri May 14 23:08:09 1999 From: jjfeiler at relief.com (John Jay Feiler) Date: Thu Nov 3 14:42:02 2005 Subject: Debugging Nightmares on OSXServer Message-ID: <199905141008.DAA29730@haque.relief.com> > Actually it's $r3 - $r10 and $f1 - $f13. And if there are mixed > integer/pointer and float arguments, the next 2 general registers (r3 - r10) > are consumed as well for placeholders of the float in the GPRs if there are > some left. But the rest was correct. Structures as arguments have > additional rules, and on the whole the rules are very complex and not worth > knowing. Oops. Sorry for the misinformation. I should have ended the comments with "I think....". Thanks to Chris for setting the record straight. If anyone really cares about exactly what is going on, you can download the PPC usage books and assembly manuals from motorola's web site as pdf documents. Speaking of which, does anyone have any hints on how to reverse engineer ObjC method from the disas output in gdb on PPC? ( not that I would ever do that, since it's obviously against half-a-dozen clauses in every software license ever written since 1973 ;-) I can do it pretty easily on an X86, but there are these weird 'addis' instructions in the PPC assembly to load the base address of the selector table into $r4 ( I think.... ) and I'm not quite sure why it is done that way.... J -- John Feiler jjfeiler@relief.com From uli at www.kaben-net.de Sat May 15 03:17:31 1999 From: uli at www.kaben-net.de (Ulrich Koester) Date: Thu Nov 3 14:42:02 2005 Subject: caught by Apples marketing Message-ID: <199905151017.MAA02487@www.kaben-net.de> Hi folks The 1999 WWDC: Now we know what Apple think about us. But first off all here is a little bit of history. In 1997 I started with Apples Yellow Box documentation. And I loved it. Fast, stable, rapid and cross plattform. Last year I started with the DR2. After a year I know what i have missed all the years. Now i know, that I am to late. This year Apple killed the Yellow Box. And we (Heino Haase Segel GmbH) have invested a lot of bucks and lot of time for the rewrite of our application. But without the cross plattform ability it is worthless. Two years for nothing. I can not believe that. For me there a two solutions. 1. Swtching to a Non-Apple development enviroment. 2. Getting a contract with Apple where there assure what they have told me two years. The second will never be happen. Where can I find GnuStep. I am so frustrated about Apple Regards Ulrich -- Ulrich Köster Heino Haase Segel GmbH Gneversdorfer Weg 9 23570 Travemünde Germany fon: +49 4502 2038 fax: +49 4502 2315 From ken_anderson at mindspring.com Sat May 15 08:11:42 1999 From: ken_anderson at mindspring.com (Ken Anderson) Date: Thu Nov 3 14:42:02 2005 Subject: caught by Apples marketing Message-ID: <000b01be9ee5$3e94ef80$340718ce@dell> Ulrich, Apple has not killed the Yellow Box. Yellow Box lives on in Mac OS X as the Cocoa libraries. However, they have stated that a strategy for Yellow/Cocoa on Windows 'has not yet been determined'. That's a lot better than 'We will not ship Cocoa on Windows'. Scott Anguish at Stepwise has asked people to write leadership@apple.com and cc: Scott so that he may put together a definitive set of reasons to stick with Cocoa on Windows. Everyone, PLEASE help Scott out here and craft a letter to Apple that explains the business reasons for them to continue on the path they promised at WWDC '97. Thanks, Ken Anderson -----Original Message----- From: Ulrich Koester To: Multiple recipients of list Date: Saturday, May 15, 1999 6:24 AM Subject: caught by Apples marketing >Hi folks > >The 1999 WWDC: > >Now we know what Apple think about us. > >But first off all here is a little bit of history. In 1997 I started >with Apples Yellow Box documentation. And I loved it. Fast, stable, >rapid and cross plattform. Last year I started with the DR2. > >After a year I know what i have missed all the years. Now i know, >that I am to late. This year Apple killed the Yellow Box. And we >(Heino Haase Segel GmbH) have invested a lot of bucks and lot of >time for the rewrite of our application. But without the cross >plattform ability it is worthless. > >Two years for nothing. I can not believe that. > >For me there a two solutions. > >1. Swtching to a Non-Apple development enviroment. > >2. Getting a contract with Apple where there assure what they have >told me two years. > >The second will never be happen. Where can I find GnuStep. > >I am so frustrated about Apple > >Regards > >Ulrich > >-- >Ulrich K?ster >Heino Haase Segel GmbH >Gneversdorfer Weg 9 >23570 Travem?nde >Germany >fon: +49 4502 2038 >fax: +49 4502 2315 > From portnoy at ai.mit.edu Sat May 15 08:19:38 1999 From: portnoy at ai.mit.edu (Stephen Peters) Date: Thu Nov 3 14:42:02 2005 Subject: Justifying the Apple Developer Connection Membership In-Reply-To: "Hunter Hillegas"'s message of "Fri, 14 May 1999 21:41:49 -0700 (PDT)" References: <19990515042752622.AAA660@nts1.webhostingprovider.com@[207.71.250.180]> Message-ID: "Hunter Hillegas" writes: > I have been a Select member of ADC for one year. I subscribed to develop for > Yellow Box. In that time, I received DR2 of Rhapsody plus the "privilege" to > buy MOSXS at a discount. That's it. $500 for that... > > [...] My membership ends 5/31. I was told by Apple that to get MOSX > PR1 I would have to re-subscribe. This is bullshit. I have gotten > next to nothing from my membership fee. I'd agree. My membership ends 6/1/99, and though I'm happier than Hunter is (I feel that the DR2 CD was nearly worth the price of admission, and I'm very happy with MOSXS), I'm overall disappointed with the way that ADC has handled YellowBox developers. I was never promised a Rose Garden, but I did feel like a third-class citizen. They won't be receiving my money again. -- Stephen L. Peters portnoy@ai.mit.edu PGP fingerprint: BFA4 D0CF 8925 08AE 0CA5 CCDD 343D 6AC6 "Poodle: The other white meat." -- Sherman, Sherman's Lagoon From rcfa at cubiculum.com Sat May 15 09:02:23 1999 From: rcfa at cubiculum.com (Ronald C.F. Antony) Date: Thu Nov 3 14:42:03 2005 Subject: caught by Apples marketing In-Reply-To: <000b01be9ee5$3e94ef80$340718ce@dell> References: <000b01be9ee5$3e94ef80$340718ce@dell> Message-ID: <199905151602.MAA29170@kannix.cubiculum.com> > Apple has not killed the Yellow Box. Yellow Box lives on in Mac OS X as the > Cocoa libraries. However, they have stated that a strategy for Yellow/Cocoa > on Windows 'has not yet been determined'. > > That's a lot better than 'We will not ship Cocoa on Windows'. Oh, come on! We *know* what that means. We knew what NeXT meant when they told us that the future of their OS "was undetermined", we know what Apple meant when they said that OSXS-intel's future "has not yet been determined". These are called EUPHEMISMS, it's a way to say "f*ck off" to people that allows the person saying it to walk off stage without getting jeers, boos and rotten eggs and tomatoes trown at him. > NeXT told us that > Scott Anguish at Stepwise has asked people to write leadership@apple.com and > cc: Scott so that he may put together a definitive set of reasons to stick > with Cocoa on Windows. Yeah, and we had written to Apple about OSXS-intel. Let me put it this way: For OSXS-intel there *was* a business case: OSXS-intel would have PRODUCED INCOME. The promised free YB-Win *only* produces costs, no income. You can search the archives, I warned people about the lack of a busines case for YB-Win for quite a while and asked the question when Apple would drop it. The second when Apple "und(et)ermined" the migration strategy YB-Win -> OSXS-intel -> OSXS-PPC and started to boast how many developers were brought back into the fold with Carbon, YB-Win was lost. YB-Win's only purpose, and the justification for the expenses Apple would incurr with it, was to be bait for developers who were about to leave the Apple platform. It was bait to make the Rhapsody strategy work. Nobody bit, because they realized there's no business case, and because they feared a move from Apple like the one we're facing right now. Apple *never* was willing to put the money where their mouth was, i.e. making an official, legally binding promise. This is once more a classical case of bait and switch. We don't need more of this apologetic talk here. When OSXS-intel was nixed, the apologists here on this list claimed: "Oh, but cross-platform isn't dead, YB-Win lives, it's just that they figured intel users would use NT anyway." Now what? Are you going to tell us "Cross-platform isn't dead, because WO-WinNT lives"? Guess what: here's another prediction: Once Apple has the support infrastructure in place and has their "pentium nuking" SMP-G4 server hardware on the market, and you can bet that's coming, you will see WO-WinNT being deemphasized. Already, the talk is how much better OSXS is doing this stuff (and it is). Soon WO-WinNT will trail in features "not possible on NT", etc. then there will be a few G4 special upgrade offers for WinNT users. They may keep a SUN/HP deployment only option for a while, since Apple doesn't have highest-end hardware, but they might even drop that, because of the load balancing possible and because NIC saturation is going to be more of a problem than CPU speed. As long as Apple thinks it's a (mainly) a hardware company, you can forget about cross-platform. Apple does not see itself like IBM as a hardware and software company that competes in each field separately. Apple bundles. Get used to it, or throw your money at GNUStep and enjoy independence. > Everyone, PLEASE help Scott out here and craft a letter to Apple that > explains the business reasons for them to continue on the path they promised > at WWDC '97. Good luck. There really are none. Apple is happy with the number of developers they have. YB/Cocoa for Win costs hundreds of thousands of dollars (if not more) in salaries for people who maintain it, and there is ZERO dollar revenue (and if there's revenue, like with the current $350/seat pricing, it's totally uninteresting a product). It doesn't serve as a migration tool either, since the critical link in the migration i.e. OSXS-intel, is missing, too. None of the big developers is going to put their head into a noose by using YB-Win for their key products, because Apple can "revise" their strategy at any point they wish (like e.g. right now). What business case? There is none! YB-Win is a goodwill product. It's a reward, a freebie, given to developers for supporting the Mac platform. It has at best a mild, long-term benefit to Apple. Apple doesn't feel like they need to reward developers for sticking with the company (despite all the broken promises). Apple is financially sound again, and with the financial health, the "screw you, if you don't like it, then leave" attitude is back. Cross-platform (at NeXT), clones (at Apple), were the right idea. The problem with both was, that the company's heart was never with it. They engaged in the policy when things got so bleak, that it was obvious what they had done wrong in the past. As soon as the money was back, the policy is dumped, because you can squeeze more money out of people if they are locked into a proprietary system. IBM had to bleed with PC clones, and only after could reap the benefits of the market explosion. MS had to give low OEM pricing of the OS first, before it could reap the benefit of having a monopoly. Apple and NeXT didn't want to or didn't feel like they were able to pay this price of admission. Sorry for being negative, but if Apple didn't see a business case in OSXS-intel, where people would have plunked down $1000/seat or where a OSXS-intel/WO bundle could have sold for $1500/CPU *and* would have gotten their OS into the enterprise *and* would have replaced a WinNT installation, then tell me where they will see a business case to spend millions of dollars over the years for developing Cocoa-Win, and give it away for free? Yeah, I don't like it a bit, but as long as things as goodwill and other non-quantifiable benefits don't figure in Apple's business decisionis, as long as the only thing that counts is quarterly profits, that long there's no case to be made for this non-product freebie. Now you may ask why they do QT for Windows, and give it away. The answer there is simple: file formats dominate the game. Plus each time QT starts, it's an Apple ad on a Windows machine. There are no Apple specific file formats or other standards that Cocoa-Win would help spread. The only people that would benefit are developers. Users see a Windows app like any other. The only time they would think of Apple is when things break: "Hm, things worked fine until I installed this Cocoa thing. Damn Apple software!" So guess what: from users Apple would only get the negative. So what's in it for Apple? Ronald ============================================================================== "The reasonable man adapts himself to the world; the unreasonable one persists in trying to adapt the world to himself. Therefore all progress depends on the unreasonable man." G.B. Shaw | rcfa@cubiculum.com | NeXT-mail welcome From rcfa at cubiculum.com Sat May 15 09:05:19 1999 From: rcfa at cubiculum.com (Ronald C.F. Antony) Date: Thu Nov 3 14:42:03 2005 Subject: caught by Apples marketing In-Reply-To: <000b01be9ee5$3e94ef80$340718ce@dell> References: <000b01be9ee5$3e94ef80$340718ce@dell> Message-ID: <199905151605.MAA29198@kannix.cubiculum.com> Oops, sorry, I thought I had posted my reply to macosx-talk. Anyway, this thread should be moved over there... Ronald ============================================================================== "The reasonable man adapts himself to the world; the unreasonable one persists in trying to adapt the world to himself. Therefore all progress depends on the unreasonable man." G.B. Shaw | rcfa@cubiculum.com | NeXT-mail welcome From mpalluat at chez.com Sat May 15 12:03:08 1999 From: mpalluat at chez.com (Marc Palluat de Besset) Date: Thu Nov 3 14:42:03 2005 Subject: caught by Apples marketing Message-ID: I was a MacOS developper, but switched to Yellow Box since Rhapsody DR1. I like Objective-C but it's nearly dead. I used to like the Yellow Box CROSS-PLATEFORM framework (Mac & Windows) but it's now dead. I don't like cocoa. I don't like Java. There is after the wwdc no more reasons to code against the cocoa APIs. It's much safer to stick with Carbon. I've worked hard on a new project since Rhapsody DR1 ... for nothing. I need now to switch to an open standard. I'm feed up with proprietary systems... and I am pissed off because Apple doesn't hold it's promesses. From malcolm at saracen.plsys.co.uk Sun May 16 05:12:44 1999 From: malcolm at saracen.plsys.co.uk (mmalcolm crawford) Date: Thu Nov 3 14:42:03 2005 Subject: caught by Apples marketing Message-ID: <199905161212.NAA00605@saracen.plsys.co.uk> Ronald wrote: > > That's a lot better than 'We will not ship Cocoa on Windows'. > Oh, come on! We *know* what that means. > As far as I can tell (and I was fortunate to have an opportunity to ask Tevanian about this personally, as well as listening carefully to others), the situation genuinely is as various folks have described -- namely that Apple is looking for a business model. First things first, let's get it out of the way: yes, this means that they reneged on assurances made over the course of the last two years, and if anyone really wanted there is apparently sufficient basis for a class action, but unless anyone really does want to go to court, shouting about broken promises is not going to make the situation any better. So: I would reiterate the suggestion that folks mail leadership@apple. Maybe start with the premise that you'd like the Cocoa/Win runtimes to be free and outline reasons why this would be a Good Thing (e.g. a wider user-base for developers). Then give a range of options for pricing that might be acceptable. $10/seat? $300/seat? $50,000/year unlimited deployment? Whatever. Maybe you wouldn't go further than having a "Made with Cocoa" splashscreen appear. In which case, fine. Tell Apple. Again as far as I can tell there is only one primary impediment to progress on this issue, and maybe he can be persuaded. One thing that I did find "interesting" was the suggestion someone (from Apple) made (and I wish I could remember who) that one reason for not shipping Cocoa/Win was that then if the same apps were available on Windows as well as MacOS (X), then there wouldn't be any advantage in having Macs. What??!! Do people really rate the OS so low that it can't be put in direct competition with Windows?! I'd love to see Cocoa apps running side-by-side on Windows and MacOS X -- nothing would better demonstrate the superior user experience and system performance of Mac OS X. mmalc. From malcolm at saracen.plsys.co.uk Sun May 16 05:18:07 1999 From: malcolm at saracen.plsys.co.uk (mmalcolm crawford) Date: Thu Nov 3 14:42:03 2005 Subject: caught by Apples marketing Message-ID: <199905161218.NAA00611@saracen.plsys.co.uk> Re Cocoa/Win runtimes: as far as I can tell (and I was fortunate to have an opportunity to ask Tevanian about this personally, as well as listening carefully to others), the situation genuinely is as various folks have described -- namely that Apple is looking for a business model. First things first, let's get it out of the way: yes, this means that they reneged on assurances made over the course of the last two years, and if anyone really wanted there is apparently sufficient basis for a class action, but unless anyone really does want to go to court, shouting about broken promises is not going to make the situation any better. So: I would reiterate the suggestion that folks mail leadership@apple. Maybe start with the premise that you'd like the Cocoa/Win runtimes to be free and outline reasons why this would be a Good Thing (e.g. a wider user-base for developers). Then give a range of options for pricing that might be acceptable. $10/seat? $300/seat? $50,000/year unlimited deployment? Whatever. Maybe you wouldn't go further than having a "Made with Cocoa" splashscreen appear. In which case, fine. Tell Apple. mmalc. From michael.maier at cityweb.de Sun May 16 05:35:59 1999 From: michael.maier at cityweb.de (Elke) Date: Thu Nov 3 14:42:03 2005 Subject: MacOS 8.6 Message-ID: <199905161235.OAA12461@mail1.cityweb.de> Hi, you will see a problem if you try to use an AppleScript in the blue box. Launch TextEdit.app and try to compile this in the AppleScript-Editor: tell app "TextEdit" end tell the GetAETE-Event doesn't work any more...and you can't compile any scripts which use YellowBox-Application Another problem: If you switch the startup disk from MacOS 8.6 to your disk image MacOS 8.5 AND you have same links to appleshare volumes or folders the system on the image will crash... In this case you have to delete the preferences called "ASLM Preferences" and restart again... regards Michael -- =========================================================== Michael Maier | | Object Factory GmbH | phone (0231)975137-0 Otto-Hahn-Strasse 18 | fax (0231)975137-99 44227 Dortmund | email m.maier@object-factory.com =========================================================== The software says Windows 98 or better, so I bought a Mac. From malcolm at saracen.plsys.co.uk Sun May 16 06:45:24 1999 From: malcolm at saracen.plsys.co.uk (mmalcolm crawford) Date: Thu Nov 3 14:42:03 2005 Subject: App layout (was Re: Finder vs Workspace) Message-ID: <199905161345.OAA00660@saracen.plsys.co.uk> On -talk Pascal wrote: > i know the openstep folks are used to the .app convention, but please give this > concept some thought, i think it's pretty compelling. > [cc-ed to -dev as Scott hasn't posted details of this yet, and I think a lot of folks missed this session which contained a number of interesting snippets.] Not, I suspect, compelling enough, given the final (rather poorly-attended) session at WWDC on Friday afternoon, in which Mike Ferris explained the new plugin architecture, and then CFBundle... ... as some may have guessed given descriptions of Core Foundation, CFBundle appears to be essentially NSBundle -- giving the same functionality including access to system resources etc -- stripped down to a C interface. It will be used by Carbon as well as Cocoa, and on MacOS 8.x as well as X; it "aims to shield details of file layout, and support multiply-locali{s,z}ed resources". This will apparently have repercussions for app layout: Apps will be organised in a hierarchy similar to the current .app layout, but with a few changes: |> A.app (gotta love those triangles... :-) ) *A (link to A executable for MacOS 8.x (+)) |> Support Files info.plist (XML property list with app-related into) |> Headers |> Frameworks |> Executables |> Mac OS 8 PPC A |> Mac OS X PPC A |> Windows A.exe |> Resources |> NonLocalizable |> Region.lproj (+ -- link to A executable for MacOS 8.x) The Finder on new versions of MacOS 8 will recognise .app folders as "executables" like MacOS X. Earlier versions, however, will see a folder. This link will be in order to give existing users faster access to a launchable file (they won't want to track down the Executables folder). Note the use of *regions* in the lproj's. These supersede the old language lproj's, allowing for support of different "versions" of a language, e.g. I guess Brazillian and Portuguese Portuguese, US and English English etc. One of the keys in the info.plist file is CFBundleDevelopmentRegion; this is used as a default region from which resources will be selected if for some reason they're absent from the user's preferred region (this provides a backstop for partial localisation -- presumably the developers at least out in all the resources for their own region!). Another issue raised in the talk was that of resource forks: apparently Carbon will provide some of this API for localised resource forks -- which will be files with resource forks and no data fork. There should be a choice of "fork or file" for Carbon, but I'm afraid it wasn't clear what the details were, and a couple of folks dominated question time with some rather pointless and unnecessarily belligerent questions. mmalc. From bbum at codefab.com Sun May 16 08:17:04 1999 From: bbum at codefab.com (Bill Bumgarner) Date: Thu Nov 3 14:42:03 2005 Subject: caught by Apples marketing In-Reply-To: Message-ID: On Sat, 15 May 1999, Marc Palluat de Besset wrote: > I was a MacOS developper, but switched to Yellow Box since Rhapsody DR1. > > I like Objective-C but it's nearly dead. I used to like the Yellow Box > CROSS-PLATEFORM framework (Mac & Windows) but it's now dead. ObjC is *not* dead. Carbon is implemented against it. ObjC *is* the language with which most of OS X, all of WebObjects, and lots of the core applications are implemented against. It *isn't* going *anywhere* anytime soon. > I don't like cocoa. I don't like Java. Cocoa == Yellow Box. They are one in the same. > There is after the wwdc no more reasons to code against the cocoa APIs. > It's much safer to stick with Carbon. In terms of cocoa, nothing has changed since last year outside of the exposure of the Quartz API and disclosure of how Classic, Yellow [Cocoa], Java, and Carbon will all live together as first class citizens. > I've worked hard on a new project since Rhapsody DR1 ... for nothing. > I need now to switch to an open standard. I'm feed up with proprietary > systems... If you want Open, switch to X w/one of the random toolkits.... ....and be prepared for serious pain. They suck compared to Carbon, Yellow/Cocoa, or other toolkits. > and I am pissed off because Apple doesn't hold it's promesses. What promises? That they would have a very cheap runtime license for Yellow? You are right-- they promised that, and didn't deliver. But they *didn't* say it was dead. The only person that implied it was Ken Bereskin... but he didn't actually say it. What was disturbing was how much Ken appears to want it and ObjC to be dead-- how much he seemed to enjoy telling folks that all their hard work was in jeopardy. Regardless, Ken doesn't make decisions about release dates or availability of Cocoa. Steve, Avie and others do. The message from those levels was not at all clear, but it was there: Mac OS X *must* be delivered Q1 of 2000. This is critical to Apple's ongoing profitability. As a result, other products-- both development and business planning-- will slip... this includes the windows environment. However, there was *no* announcement of the Windows Yellow environment being dead. As a matter of fact, Cocoa on Windows must continue to exist-- WebObjects depends on it and there is to be a release of 4.5 of WebObjects this year. b.bum From gerard.iglesias at ina.fr Sun May 16 08:26:07 1999 From: gerard.iglesias at ina.fr (Gerard Iglesias) Date: Thu Nov 3 14:42:03 2005 Subject: caught by Apples marketing References: <199905161212.NAA00605@saracen.plsys.co.uk> Message-ID: <373EE38F.DC8936AA@ina.fr> mmalcolm crawford wrote: > What??!! Do people really rate the OS so > low that it can't be put in direct competition with Windows?! I'd > love to see Cocoa apps running side-by-side on Windows and MacOS X -- > nothing would better demonstrate the superior user experience and > system performance of Mac OS X. It is the most important point of the all story. I am convinced of this. Suppose that you have to speak with your manager about a new software very useful that run under a very wonderul hardware, and your company don't own this wonderful hardware/software product. two case: 1/ This software would bring a lot of benefit to your company, then they decide to buy the complete package. 2/ This software would bring a lot of benefit to your company but a version exist on the dummy, boring harware/OS that your company own in quantity. Then the company will by the software version that run under this harware. It is simple, and obvious if you live in a company who own a majority of this very nad hardware/software association. Apple have to face to this situation, and their solution is to bring powerful solution at the right price, like Final Cut Pro for example on their hardware only. Or if in the near future the power of the PowerPC is so big, and the quantity of Macintosh sold so good, that the decision to be completely cross-platform will not inluence too much their business model hardware oriented. Think about Microsoft selling office on PC and Macintosh, obviously a less quantity of Windows is sold, but not enough to be in danger. We will not see a version of Final Cut Pro on Intel before a very very long time, unless never. Sincerely. Gerard From buck.erik at mcleod.net Sun May 16 12:05:26 1999 From: buck.erik at mcleod.net (Erik M. Buck) Date: Thu Nov 3 14:42:03 2005 Subject: screwed by Apples marketing/deceit Message-ID: <000201be9fcf$83a30c20$762310d0@elcry> I left this list last Jan or Feb because I deemed it a waste of time and nobody else seemed to agree with me that Apple was/is actively courting its own destruction through broken promises, broken contracts, outright deceit, and a complete lack of clues. Actually a few people did agree with me, but they all left too. The bottom line is that nobody (statistically speaking) cares about Apple only technology. Apple has taken the best development environment available for platforms that people do (statistically speaking) care about, and they have killed it. It does not matter if it lives on as Apple only technology. For the rest of the world it is dead. Apple's actions compounded with their outright lies and deceit have hurt a lot of people financially. I know that Apple says WebObjects will continue cross platform. Should we believe them ? They said Yellow Box was their cross platform API. Ironically, I do not really care about YellowBox or Cocoa. All I really need is Openstep Enterprise 4.2 deployment licenses at a reasonable price like the promised less that $20/seat price. I have contractual obligations for deployment that date from a time when NeXT actively sold and promoted the technology at a price of $500 for one developer seat and 5 deployment seats. That comes to less than $100 per seat and could be considered free deployment to 5 seats because you are paying for the developer seat. Let them move on as an Apple only technology with Cocoa for all I care. Sell the obsolete Openstep Enterprise 4.2 to the corporations that frankly will never ever buy hardware from Apple again. Are all of the people remaining on this list truly comfortable with the prospect of developing software for Apple only and linking their fate with the caprices of Apple ? For the past 10 years companies have spent billions to get away from Apple. Are you all convinced that those companies were silly to open their options. Is the Yellow/Cocoa technology so compelling that you are willing to give up access to 96% of the potential market and 99.8 % of the corporate vertical app market in order to use it ? I could sell Next technology into corporations. I can not and will not try to sell Apple hardware to my Fortune 100 customers. Scott Anguish made an interesting comment that he needs to deploy an Openstep/YellowBox application and that he has no way to do it. He is not the only one in that position. My company has an application that we successfully deployed for years before Apple pulled the rug out from under us. One year ago I complained that I could no longer deploy because Apple was unwilling to sell deployment licenses for Openstep Enterprise 4.2 at any price. The replies I got (From Apple) included a variety of prices for deployment ranging from (It is not a product) to $1,500 per seat to $300 per seat for deployment. Given no other choice, I have been forced to live with those prices. Other vendors including Omni Development and Scott Anguish are presumably also unable to deploy effectively on Windows (I can only assume they have the same contacts and contracts with Apple that we have.) I just do not understand how people at Apple can continue to work for that company and maintain any self respect. Specifically, If I was in Ernest Prahabaker's (sp?) shoes, I would have quit a long time ago. Being charitable, Apple has made him a liar. Being less charitable, he has no honor himself. In either case he should not remain with Apple. Why do I pick on him ? He is one of the people who made undeniable promises (probably with the best intentions) on which he and Apple have reneged. Of course Jobs himself said the same things so how can really blame Ernie ? From sanguish at digifix.com Sun May 16 14:49:03 1999 From: sanguish at digifix.com (Scott Anguish) Date: Thu Nov 3 14:42:03 2005 Subject: Mac OS X questions... [namely, when do us non-attendees get In-Reply-To: <1285396912-5321042@evatac.com> References: <1285396912-5321042@evatac.com> Message-ID: <199905162149.RAA08350@digifix.com> Tom Carstensen wrote: > > > >I haven't heard, but in past years developer program members have > >received all (or almost all) the WWDC "goodies" within a month of the > >conference, usually in the next month's developer mailing. (The one > >exception was two years ago, when MacOS developers didn't get the copy of > >WebObjects that was distributed at WWDC.) > > I'm personally upset that WWDC attendees got it and we non-attendees > did not. > Funny, but I'd be upset (as one who paid $$$ to go to WWDC) if non-attendees got all the stuff that we did.. But hey.. > A month??? uh.. putting up on the developer ftp will take all of > a few minutes. Personally i can do totally with the monthly cd > mailing -- i get it when it's announced via ftp. > Does nobody comprehend that a 300Mb download of a CD image just isn't viable? This ain't Mac OS. You need to have a full CD to start the install from. > There is *clearly* some discrimination here, has it's just as > easy to ftp it as it is to hand it out at the conference. > Nope. Not in any way, shape, or form is it as easy. And lets not even TALK about the piracy/moral flexibility that some Apple devs have about passing this stuff around. > I agree -- what do we pay $500/yr for? We are less of developers > becuase apple didn't get our $1000 for the WWDC? > No, but maybe, just maybe, those who paid to get the stuff at the conference should get the first crack. > -Tom Carstensen > > > > -- > Tom Carstensen tom@evatac.com > President Evatac Enterprises 877.287.8618 (toll free) > > Macintosh, Windows & Be Software Development > From sanguish at digifix.com Sun May 16 15:01:51 1999 From: sanguish at digifix.com (Scott Anguish) Date: Thu Nov 3 14:42:03 2005 Subject: death of YB/Windows In-Reply-To: <9905122340.AA12369@vviuh221.vvi.com> References: <9905122340.AA12369@vviuh221.vvi.com> Message-ID: <199905162201.SAA08408@digifix.com> David Young wrote: > > -- At this time, we have no formal commitment to Cocoa on any > > platform other than Mac OS X Server > > Probably another Ken Bereskin foot in mouth. I suspect he meant or said > "Mac OS X and Mac OS X Server", i.e. not just Server. Did I actually write Mac OS X Server??? (Checks notes and site) Yes, I did... I suppose it is possible that Ken said this with the Server intact, but I don't know for sure.. This could have been a slip on my part. I guess only those with the tapes will know for sure. I'd like to know for sure, since I already suspect Ken Bereskin hates me, and I'd definately clarify if it was just my slip here.. Steve said during the keynote that it (cocoa) was an integral part of the OS X product (with both Objective-C and Java APIs.. he said that specifically). The fact that __nobody__ pointed this Server comment out to me kinda surprises me. Two Apple folks commented on the statements that had immediately followed it about the "potential uncertainty" of WebObjects on NT's future (which lead to Allen Denison's posted statement). I further got two good pieces of feedback from Chris Kane about some potential areas of confusion in the CoreFoundation and Finder notes (which I hope to get up on the site tonight). Cocoa (which is MUCH better to eat than a "Yellow Box" or Carbon for that matter) is secure in its future on Mac OS X... the question is about the licensing on Windows. There should be no question about this, Apple had made a commitment.. From kcd at jumpgate.com Sun May 16 15:11:36 1999 From: kcd at jumpgate.com (Kenneth C. Dyke) Date: Thu Nov 3 14:42:03 2005 Subject: WWDC notes & spoilers Message-ID: <199905162206.PAA10322@vorlon.jumpgate.com> > From: "Mark F. Murphy" > Date: 1999-05-16 07:32:16 -0700 > To: Multiple recipients of list > Subject: Re: WWDC notes & spoilers > X-From_: server@omnigroup.com Sun May 16 07:24:52 1999 > X-Listprocessor-Version: 6.0 -- ListProcessor by Anastasios Kotsikonas > Originator: macosx-talk@omnigroup.com > X-Comment: To unsubscribe, follow directions at > http://www.omnigroup.com/MailArchive/ > > At 4:38 PM -0700 5/11/99, Tom Hageman wrote: > >Apple's been there, done that... or at least attempted to "modernize" > >ObjC somewhere in '97. > > No they didn't. > > They floated the idea. > > >Was quickly put down due to stolid resistance by > >some of us old ObjC curmudgeons. (Hey, if it ain't broke, don't fix it. > >If you really want C++ syntax, use Java ;-) > > So, keep using that syntax. Nothing would force you to change. Well....that's not entirely true. One of the things *both* groups of programmers would have to start doing is providing ways to map back and forth between selector names between the two different syntaxes, just as we have to do work to wrap our ObjC code in Java right now. Otherwise someone using the "new" syntax might not be able to use someone else's ObjC classes, and vice-versa. Obviously this would not be a good thing for the community as a whole. > What I find amazing is how some are opposed to having two syntax sets for > the functionality of Obj-C. > > What are you all scared of? Well, I can think of cases where attempting to fully use the C++/Java syntax would actually break certain things. For example, in ObjC I can do this: (Interface skipped to save space) typedef void (*func_ptr)(void); @implementation MyClass : NSObject { func_ptr someInstanceVariableName; } - (func_ptr)someInstanceVariableName { return someInstanceVariableName; // Return value of instance variable. } + (func_ptr)someInstanceVariableName { return NULL; } - (func_ptr)someOtherMethod { someInstanceVariableName(); // No question at all what this is supposed to do. return [self someInstanceVariableName]; // No question as to what this does. } @end; Now, note that these two usages never conflict because there is a syntactical difference between C contstructs and invoking methods. There is no problem at all between allowing method names to have the same name as an instance variable. Now if we try to use a C++/Java-like syntax, we suddenly run into this problem: class MyClass : public NSObject { func_ptr someInstanceVariableName; virtual func_ptr someInstanceVariableName(void); static func_ptr someInstanceVariableName(void); virtual void someOtherMethod(void); }; func_ptr MyClass::someInstanceVariableName(void) { return someInstanceVariableName; // Does this return the value of the instance variable, or a pointer to the method by the same name? } static func_ptr MyClass:someInstanceVariableName(void) { someInstanceVariableName(); // What does this do?? Call a class method? Instance method? Use the function pointer? return someInstanceVariableName; } func_ptr MyClass::someOtherMethod(void) { someInstanceVariableName(); // What does this do?? Call a class method? Instance method? Use the function pointer? return someInstanceVariableName(); } Somehow I bet there isn't a single C++ compiler that would allow you to have 3 different members with the same name, mostly because there is no syntactical way to tell them apart by usage. Obviously this is somewhat of a pathological case, but ObjC has no problems with this sort of thing specifically because the language designers did not try to re-use the C syntax everywhere possible. You could start putting all kinds of non-Java/non-C++ restrictions. For example, when making method calls you must always use "self" or "this". Of course again, even with a pointer to MyClass, what would this do: x = myClassInstance->someOtherMethod(); This could still have two entirely different meanings with the C++/Java syntax (either access the ivar directly, or call the method by the same name). This just becomes one really big huge mess in a hurry, IMHO. The beauty of ObjC in my opinion is that it *doesn't* try to overload C's syntax to mean different things. You always know just by looking at the code *exactly* what the semantics of your statements are. We don't have to have C++'s list of disabiguation rules to figure things out. ;) -Ken From rcfa at cubiculum.com Sun May 16 14:27:07 1999 From: rcfa at cubiculum.com (Ronald C.F. Antony) Date: Thu Nov 3 14:42:03 2005 Subject: App layout (was Re: Finder vs Workspace) In-Reply-To: <199905161345.OAA00660@saracen.plsys.co.uk> References: <199905161345.OAA00660@saracen.plsys.co.uk> Message-ID: <199905162127.RAA04748@kannix.cubiculum.com> [mmalc's description snipped] Excellent news! Ronald ============================================================================== "The reasonable man adapts himself to the world; the unreasonable one persists in trying to adapt the world to himself. Therefore all progress depends on the unreasonable man." G.B. Shaw | rcfa@cubiculum.com | NeXT-mail welcome From rcfa at cubiculum.com Sun May 16 14:46:05 1999 From: rcfa at cubiculum.com (Ronald C.F. Antony) Date: Thu Nov 3 14:42:03 2005 Subject: caught by Apples marketing In-Reply-To: <199905161212.NAA00605@saracen.plsys.co.uk> References: <199905161212.NAA00605@saracen.plsys.co.uk> Message-ID: <199905162146.RAA04810@kannix.cubiculum.com> > One thing that I did find "interesting" was the suggestion someone > (from Apple) made (and I wish I could remember who) that one reason > for not shipping Cocoa/Win was that then if the same apps were > available on Windows as well as MacOS (X), then there wouldn't be any > advantage in having Macs. What??!! Do people really rate the OS so > low that it can't be put in direct competition with Windows?! I'd > love to see Cocoa apps running side-by-side on Windows and MacOS X -- > nothing would better demonstrate the superior user experience and > system performance of Mac OS X. Well, the key thing to note, and that's what the person at Apple probably meant, is that people are not operating in the "homo oeconomicus" model of rational decision making based on *facts*, but they are rather irrational and decide based on what is *perceived to be the case*. So if applications are available on both the Mac and on Windows, but there are still more apps on Windows total, then people will *perceive* no difference between the two, like they did, after the introduction of Win3.* *perceive* that there was no longer a significant advantage in the UE between a Mac and Windows, and bought Windows en masse. I fully agree, that *if* people could/would compare OSX and Win* side by side, OSX would win hands down. But people don't go through that effort, even less so since *due to the cancellation of OSX-intel* that entails plunking at least a thousand bucks for new hardware onto the table. Which brings us back to why OSX-intel is crucial, but we have beaten that horse into a pulp and sent plenty of messages with rather decent arguments to leadership@apple.com w/o useful result. It's the same old problem that NeXTSTEP had: *if* more people had known about it, and had used it, it might have become a success. Only too late in the game was realized that even the smallest segment of the intel market still is bigger than the market NeXT hardware alone can carve out, and hence cross-platform was born, but too late. Same as Mac clones came too late. Now that Apple essentially has an entirely new platform, yet backwards compatible, when they could do the cross-platform stuff early in the game, when it's useful, they forget the old lessons. Maybe in a few years from now, when MS has comparable technologies, and starts gnawing at the Mac marketshare again, then they will decide otherwise, but then it's too late. Apple needs to buy market share, and that means lower short term profits. As long as they are not willing to take that risk, they will never get there. MS on the other hand is doing it constantly, they pay for things that don't bring them a cent of profits directly, only to hurt competition and spread their technologies' reach, well knowing that eventually this will pay off through a larger market share. Apple doesn't want to go there. They take refuge in whatever niche MS leaves them... There is no business case to be made, as long as that means profits next quarter. If they are talking profits 5 years from now, then maybe, but obviously that's not their operating mode, because otherwise they would also consider what message this continued promise breaking sends to enterprise customers. It says: "Use Apple technologies right now, while they are leading, and while it can get you a quick ROI. But even while your're developing the short term, Apple-based solution, keep in mind reengineering the software with tools from a more reliable source." And if you look at how many of WO's show-case clients have switched to other solutions, that is interesting. (Yes, there was not always fair-play in these situations, BUT nobody dumps a winning horse if they don't have at least some doubts about the long-term prospects.) Ronald ============================================================================== "The reasonable man adapts himself to the world; the unreasonable one persists in trying to adapt the world to himself. Therefore all progress depends on the unreasonable man." G.B. Shaw | rcfa@cubiculum.com | NeXT-mail welcome From rcfa at cubiculum.com Sun May 16 14:45:56 1999 From: rcfa at cubiculum.com (Ronald C.F. Antony) Date: Thu Nov 3 14:42:03 2005 Subject: screwed by Apples marketing/deceit In-Reply-To: <000201be9fcf$83a30c20$762310d0@elcry> References: <000201be9fcf$83a30c20$762310d0@elcry> Message-ID: <199905162146.RAA04804@kannix.cubiculum.com> > I just do not understand how people at Apple can continue to work for that > company and maintain any self respect. Specifically, If I was in Ernest > Prahabaker's (sp?) shoes, I would have quit a long time ago. Being > charitable, Apple has made him a liar. Being less charitable, he has no > honor himself. In either case he should not remain with Apple. Why do I > pick on him ? He is one of the people who made undeniable promises > (probably with the best intentions) on which he and Apple have reneged. Of > course Jobs himself said the same things so how can really blame Ernie ? While I agree with most of what you said, I would most certainly *not* blame Ernie. He has been again, and again one of the most helpful people of the bunch, constantly listening to customer's requests and concerns. He is also big time involved with opening up Apple in many areas, Darwin included. It would absolutely the *worst* if he left. We need *more* Ernie's at Apple. If you want to get rid of people, start with Ken Bereskin, who talks all of Apple's interesting cross-platform products so long into oblivion until the lack of demand resulting from it, turns it into a self-fulfilling prophecy. (Then again, that's his job, to get the bad news out, and take the flak. Who knows...). In any case, the responsibility is to be found higher up in the food chain. We can't go on just turning Jobs into a hero for the iMac (which probably saw a good deal of development already under Amelio), while being blind to his role in connection of OSX-intel, YB-Win and soon (if the trend continues) WO-Win. Ronald ============================================================================== "The reasonable man adapts himself to the world; the unreasonable one persists in trying to adapt the world to himself. Therefore all progress depends on the unreasonable man." G.B. Shaw | rcfa@cubiculum.com | NeXT-mail welcome From jmederos at gran-via.com Sun May 16 16:13:48 1999 From: jmederos at gran-via.com (Jorge Mederos) Date: Thu Nov 3 14:42:03 2005 Subject: screwed by Apples marketing/deceit References: <000201be9fcf$83a30c20$762310d0@elcry> Message-ID: <373F512C.48C89B2C@gran-via.com> Erik M. Buck wrote: > > The bottom line is that nobody (statistically speaking) cares about Apple > only technology. At least in the enterprise market this is true. The only light on the enterprise horizont is WebObjects. Why? because it's from Apple? No. Because it's good and it's cross-platform. It's not very difficult to get people receptive to WebObjects because it's deployed on Sun or HP servers, developed (at least until now) on NT or MXS, supports all important DB, CORBA and can be written in Java or Obj-C. If WebObjects moves, or if there is a risk or fear that WebObjects may move towards a more propietary Apple hardware, much of its good mindshare will be lost. I only see basic important support for current Apple's enterprise customers, but (as Erik and many others have said here) I don't see any good migration path for them. All they have is a broken path after OpenStep 4.2 with the demise of MXS intel. 4.2 was a good start but now it's not anything beyond. Bad thing is that. If Apple wish to have any possibility in the enterprise market in few years, they must begin now. They must begin repairing broken paths. If they only care about selling iMacs to secretaries... ok, they only need to care about Microsoft supporting Office on MacOS X. If there is not Office, there wont be iMacs in the enterprise either. What enterprises cares about are Long-Term Reliable Strategies (LTRS). It's also applicable to iMacs with Office. Enterprises assume a lot of risks in their own business, and there is no need to assume more selecting a propietary hardware platform that seems to become more propietary in time. Breaking past declared strategies is undermining this basic principle. > I know that Apple says WebObjects will continue cross platform. Should we > believe them ? They said Yellow Box was their cross platform API. I think this is the kind of bad mindshare Apple is obtaining for changing their focus to a more propietary platform breaking past strategies. It doesn't matter that WO wont ever lost its cross-platform capabilities, if there is a fear that it would be. LTRS apply everywhere in the enterprise. Recent moves with DO don't help very much with this issue. > Ironically, I do not really care about YellowBox or Cocoa. All I really > need is Openstep Enterprise 4.2 deployment licenses at a reasonable price > like the promised less that $20/seat price. I have contractual obligations > for deployment that date from a time when NeXT actively sold and promoted > the technology at a price of $500 for one developer seat and 5 deployment > seats. That comes to less than $100 per seat and could be considered free > deployment to 5 seats because you are paying for the developer seat. Let > them move on as an Apple only technology with Cocoa for all I care. Sell > the obsolete Openstep Enterprise 4.2 to the corporations that frankly will > never ever buy hardware from Apple again. What it's more ironic is that all enterprise providers of now-Apple technologies are not dissapointed with the technology, but with the marketing strategies. They only need a path for the future. What path? One that continues past strategies. If Apple doesn't want to sell YB-NT (oops! Cocoa) for the masses yet... well, then there should be an alternative. Like selling the actual YB implementation as Apple's OpenStep 5.0 with higher prices. This makes things a little more difficult, but there is a chance that good folks that had bet and carefully listened to Apple will survive. After all, OpenStep always was targeted to vertical markets. The path may be slightly different not reaching the masses, but it should not be broken. If Apple doesn't want to loose actual enterprise customers they should sell a limited version of MXS for intel, calling it for example Mac OS X Enterprise Server and being sure that it works on most important PC configurations: Dell, HP, IBM, Gateway and a few more. With the public Darwin base, it should not be difficult to maintain a limited set of drivers for it. Be consistent, be reliable, call the actual implementation of YB OpenStep 5.0 or 5.1 and sell it to the enterprise market. Putting higher prices on the OS is the way to limit its adoption, if Apple doesn't want to appear as an Intel guy. Sell the OS as a unix-class server. Don't put all the easy of use stuff that always will be standard on Mac hardware, like netbooting iMacs or the BB (Classic) environment. Be sure that samba works correctly on it and try to unify code bases for the OS on mach 3.0. The Darwin open source initiative is a good LTRS. Work with it on a product that also gives money, customers and good mindshare. One of the most interesting parts of MXS is that many open-source software is available for it... so it's not closed. I can use the same build and test environments that I'm currently using with our Sun and HP servers. The code base for cross-platform Unix-like programs is similar. This is reliable, this is what I expect from a hardware/software company. It's a LTRS. Nobody wants potentially broken paths. With good LTRS it's possible to sell Apple everywhere. Apple has a good LTRS in the consumer market, so people feel ok buying iMacs. Why don't do the same with the enterprise market? WO does something that is strange with Apple in general, yet does it surprisely well. WO is cross-platform and support a lot of third-party tools and environments, thing that make easy to be adopted by many customers. Regards, Jorge. From mtrent at msn.fullfeed.com Sun May 16 16:35:11 1999 From: mtrent at msn.fullfeed.com (Mike Trent) Date: Thu Nov 3 14:42:03 2005 Subject: screwed by Apples marketing/deceit Message-ID: <199905162335.SAA00872@fullfeed.msn.fullfeed.com> This is a perfect example of a message that doesn't belong in macosx-dev. No matter how I try, I can't see how personal attacks on Ernie, Ken, and Steve are appropriate here. Take it to talk or see if Omni will create macosx-advocacy, macosx-screwed, or some other mailing list. Thanks, Mike Trent ---------- >From: "Ronald C.F. Antony" >To: Multiple recipients of list >Subject: Re: screwed by Apples marketing/deceit >Date: Sun, May 16, 1999, 4:15 PM > >> I just do not understand how people at Apple can continue to work for that >> company and maintain any self respect. Specifically, If I was in Ernest >> Prahabaker's (sp?) shoes, I would have quit a long time ago. Being >> charitable, Apple has made him a liar. Being less charitable, he has no >> honor himself. In either case he should not remain with Apple. Why do I >> pick on him ? He is one of the people who made undeniable promises >> (probably with the best intentions) on which he and Apple have reneged. Of >> course Jobs himself said the same things so how can really blame Ernie ? > > While I agree with most of what you said, I would most certainly *not* > blame Ernie. He has been again, and again one of the most helpful people > of the bunch, constantly listening to customer's requests and concerns. > He is also big time involved with opening up Apple in many areas, Darwin > included. It would absolutely the *worst* if he left. We need *more* > Ernie's at Apple. > If you want to get rid of people, start with Ken Bereskin, who talks > all of Apple's interesting cross-platform products so long into oblivion > until the lack of demand resulting from it, turns it into a self-fulfilling > prophecy. (Then again, that's his job, to get the bad news out, and > take the flak. Who knows...). > In any case, the responsibility is to be found higher up in the food > chain. We can't go on just turning Jobs into a hero for the iMac (which > probably saw a good deal of development already under Amelio), while > being blind to his role in connection of OSX-intel, YB-Win and soon > (if the trend continues) WO-Win. > > Ronald > ============================================================================== > "The reasonable man adapts himself to the world; the unreasonable one persists > in trying to adapt the world to himself. Therefore all progress depends on the > unreasonable man." G.B. Shaw | rcfa@cubiculum.com | NeXT-mail welcome > > From buck.erik at mcleod.net Sun May 16 17:29:20 1999 From: buck.erik at mcleod.net (Erik M. Buck) Date: Thu Nov 3 14:42:03 2005 Subject: screwed by Apples marketing/deceit Message-ID: <002201be9ffc$4ed35400$5b2b10d0@elcry> I believe this topic belongs on this list because this topic seeks an answer to a partially technical question: Is Cocoa better enough to justify sacrificing 96-99.8 % of the market for our applications ? To answer the question we must discuss technology and business simultaneously. Technically speaking, Cocoa technology is without peer. Products like IBM's Visual Age and the former Delphi are good, but not quite as good for a variety of reasons. [Another discussion perhaps]. Nobody here really contests the virtues of the technology. I seek to quantify that virtue if possible and contrast it with the non technical limitations such and limited deployment to 4% of the market possibly beginning 1 year from now. For instance, the"new" Interface Builder supports Active-X Controls and Java Beans. The existing IB components such as NSTextView compete with the best Active-X controls and Java Beans. Now we could hypothetically have the best of both. How much is that worth to us ? Active-X controls are not available for the only deployment platform. Oops --- I guess we can not count those in our technical evaluation... You see where I am going right ? From rcfa at cubiculum.com Sun May 16 21:12:21 1999 From: rcfa at cubiculum.com (Ronald C.F. Antony) Date: Thu Nov 3 14:42:03 2005 Subject: screwed by Apples marketing/deceit In-Reply-To: <199905162335.SAA00872@fullfeed.msn.fullfeed.com> References: <199905162335.SAA00872@fullfeed.msn.fullfeed.com> Message-ID: <199905170412.AAA06021@kannix.cubiculum.com> mtrent@msn.fullfeed.com wrote: > This is a perfect example of a message that doesn't belong in macosx-dev. No > matter how I try, I can't see how personal attacks on Ernie, Ken, and Steve > are appropriate here. Take it to talk or see if Omni will create > macosx-advocacy, macosx-screwed, or some other mailing list. [rest snipped] Dude! Have you ever gotten so much mail that you slipped to check on a particular message after hitting the "reply" button if the address filled in automatically is actually the forum for which a particular discussion is appropriate for? I generally try to take care of redirecting replies to the proper group, but with the general mail volume I keep up with, such accidents happen. How about you chill out and leave the job of being list-mom to the people who actually *are* list hosts, rather than stepping in and assuming a role that nobody asked you to take on? In particular: if you're so concerned about off-topic posts, how about you send a nice and polite message to the persons in question *privately*, rather than wasting everyone's time by launching a new off-topic thread on off-topic posts? That's about as silly as the people who reply to the "how do I sign off this list" posts with answers that "such questions shouldn't be sent to the list, but to the list owner, and that generally RTFM" instead of either simply ignoring the question, or sending directly a helpful response. Ronald ============================================================================== "The reasonable man adapts himself to the world; the unreasonable one persists in trying to adapt the world to himself. Therefore all progress depends on the unreasonable man." G.B. Shaw | rcfa@cubiculum.com | NeXT-mail welcome From uli at www.kaben-net.de Mon May 17 00:44:29 1999 From: uli at www.kaben-net.de (Ulrich Koester) Date: Thu Nov 3 14:42:03 2005 Subject: screwed by Apples marketing/deceit Message-ID: <199905170743.JAA05627@www.kaben-net.de> A non-text attachment was scrubbed... Name: not available Type: text/enriched Size: 1520 bytes Desc: not available Url : /mailman/archive/macosx-dev/attachments/19990517/fb831dfc/attachment.bin From lavoie at cst.ca Mon May 17 06:18:48 1999 From: lavoie at cst.ca (Martin-Gilles) Date: Thu Nov 3 14:42:03 2005 Subject: caught by Apples marketing Message-ID: <199905171318.JAA16612@plexus.cst.ca> >Apple has not killed the Yellow Box. Yellow Box lives on in Mac OS X as the >Cocoa libraries. However, they have stated that a strategy for Yellow/Cocoa >on Windows 'has not yet been determined'. > >That's a lot better than 'We will not ship Cocoa on Windows'. Wake up. Apple said the same thing for Rhaptel. It's letting developers down "easy". =========================================================================== Martin-Gilles Lavoie, Mac OS programmer, Corporate Software & Technologies --------------------------------------------------------------------------- (internet) http://pages.infinit.net/mouser/ mouser@videotron.ca (intranet) http://blackhole.cst.ca/ lavoie@cst.ca --------------------------------------------------------------------------- "No! Try not. Do, or do not. There is no try." -- Yoda on error handling From bbum at codefab.com Mon May 17 06:24:56 1999 From: bbum at codefab.com (Bill Bumgarner) Date: Thu Nov 3 14:42:03 2005 Subject: screwed by Apples marketing/deceit In-Reply-To: <199905170743.JAA05627@www.kaben-net.de> Message-ID: (there is a -dev component to this) On Mon, 17 May 1999, Ulrich Koester wrote: > The YellowBox is dead. It is *not* dead. Apple made *no* announcement that they are killing any cross platform development solution. None. I repeat. Yellow as a cross platform solution is *not* dead. It has been renamed. It is now called Cocoa. This is the same thing. Apple did fail to announce any kind of licensing scheme or deployment plan. Nor did they announce time frames. Their message was unclear and went against committments made in the past. That was bad. However, they did *not* kill the product. It is *not* on the same list as OpenDoc, Pink, AUX, or Newton. Apple is a hardware company. For a cross platform deployment solution to make sense, that solution must help sell Mac hardware or otherwise generate significant revenue. Every effort the company is expending appears to be to those ends; it is critical that OS X ship in a timely manner (Q1 next year) and that WebObjects remains on a sesible development/release cycle. To these ends, a Yellow cross platform solution is not a critic piece. The replacement of the imaging layer [Quartz replacing Display Postscript] will require siginificant engineering to make it work on the Windows platform while contributing nothing to the WebObjects release schedule in the near future. However, consider this (here we get to the -dev part): Mac OS X is an XML based system. Anywhere a property list is used now, XML can apparently be used instead. This includes InterfaceBuilder; XML is used to describe NIB contents. By decoupling the interface description from any kind of an Object Stream, it means that NIB files can easily be used from Carbon/Cocoa/Java applications or other interface consuming technologies. Very Cool. Also: Mac OS X treats PDF as a first class citizen. Carbon/Cocoa apps can easily consume, render and produce PDF. The print panel can "save as PDF". You can drag PDF docs into a Text view and it "just works". Very Cool, part two. Now, given the above, the evolution of the Web marketplace, and given NeXT's (and, now, Apple's) long history of maintaining consistent APIs across product lines (even across languages; the core foundation in Carbon/Cocoa basically presents Foundation functionality to either an ObjC or ANSI-C API.... transparent to developer), I would conjecture that: - Apple will move Quartz/Cocoa (Cocoa = (Yellow - DPS) + Quartz) to the Windows platform. At WWDC, it was discussed several times that Development and Deployment of WebObjects will remain a first class citizen (i.e. supported and evolved with the rest of the platform) on Windows. This makes sense; a WebObjects development environment with transparent handling of XML and PDF is going to be a hell of a selling point. - The move to Windows will require a significant engineering effort-- Windows display management is hard-- but will result in a much faster/more usable system. Display PostScript is quite resource intensive and was designed for a systme that effeciently messages and moves memory between apps. Windows does neither and DPS based apps suffer accordingly. Regardless, it will take time to move Quartz to the new platform-- Time and engineering effort Apple cannot expend if it wants to make the OS X release date. Now, combine all of the above with the complete shambles that is the Windows market. To what API should the quartz engineering be done against??? The existing Windows NT/98 split APIs-- building the light weight window management and display infrastructure will require two sets of significant engineering. Or do you build it to Win2000 and hope that that product ships in a reasonable time frame (it hasn't yet-- it was supposed to ship Q1 of '98) and remains stable for any length of time? None of this is to say that Apple won't kill the windows Cocoa *deployment* solution. IMHO: I don't think they will, but I am confident that the availability of such will not happen until OS X ships and Apple has a compelling Apple Hardware based environment that works 10X better. b.bum Some Dev Questions: - Anyone have example code that shows use of the same core foundation structure / object from both ANSI C and ObjC? - Can you stick an EIDE drive in a Beige computer along with the U/W SCSI drives and boot Mac OS X off of the EIDE, Mac OS X Server off of hte U/W SCSI card? From thomas.roehl at cai.com Mon May 17 06:49:49 1999 From: thomas.roehl at cai.com (Roehl, Thomas) Date: Thu Nov 3 14:42:03 2005 Subject: Mac OS X questions... [namely, when do us non-attendees get Message-ID: <4C094419B0E6D2119FF100805F85CABE72F174@usilmse1.cai.com> >> I agree -- what do we pay $500/yr for? We are less of developers >> becuase apple didn't get our $1000 for the WWDC? >> > No, but maybe, just maybe, those who paid to get the stuff at the > conference should get the first crack. There is nothing wrong with that at all. But Apple should at least let it's registered developers who didn't go to WWDC know what kind of access they will get to it. I joined the developer program to get early access to MacOS X client/server and the program IMHO really does not do that. I personally feel that Apple has not acted in good faith in representing what developers will get access to by joining the developer programs. From dyoung at vviuh221.vvi.com Mon May 17 07:46:06 1999 From: dyoung at vviuh221.vvi.com (David Young) Date: Thu Nov 3 14:42:03 2005 Subject: DPS && Windows not slow. References: Message-ID: <9905171446.AA18395@vviuh221.vvi.com> > Display PostScript is quite resource intensive and was > designed for a systme that effeciently messages and moves memory > between apps. Windows does neither and DPS based apps suffer accordingly. Not true. Our validation tests show that in many real-world examples, with identical YB source code, that drawing on Windows using DPS is at least as fast, perhaps about 2% faster, as compared with OpenStep/Intel on identical Intel hardware. In fact, because of our drawing algorithms, many of our YB-based features (Cocoa-based) are actually faster than others' native WIN32 applications. ************************************* By the way; to summarize the current state of YB: YellowBox, the code name, is renamed Cocoa, the product name. Cocoa IS on Mac OS X ; and Mac OS X Server now. Objective-C is still on Mac OS X * ; but is not considered a marketing buzzword so it's not included in front-line hype. YellowBox IS available for sale on Windows at decent Systems Integrators' prices, but not ISV prices. YellowBox's Foundation is THE foundation of WebObjects, the most popular app server on the web. Apple decided that YellowBox was valuable and shouldn't be given away for near-free; that enterprise customers who want it should have to actually pay for it. etc., etc. Sounds reasonable. ************************************* Also, I'm a collector of history. I cached this html from Apple's web site: http://www.vvi.com/~dyoung/yellowbox.html knowing that history would be erased in the volatility of the web. It is, of course, for my own personal use only. ************************************* Thanks A Bunch! David Young; VVI-DCS dyoung@vvi.com From lavoie at cst.ca Mon May 17 07:43:10 1999 From: lavoie at cst.ca (Martin-Gilles) Date: Thu Nov 3 14:42:03 2005 Subject: DPS && Windows not slow. Message-ID: <199905171443.KAA17340@plexus.cst.ca> >Apple decided that YellowBox was valuable and shouldn't be given away for >near-free; that enterprise customers who want it should have to actually pay >for it. etc., etc. Sounds reasonable. Try saying reasonable to those that have been waiting to deploy on winblows. Hints: Wear a helmet. Be in a padded cell. Be affraid one finds the key. =========================================================================== Martin-Gilles Lavoie, Mac OS programmer, Corporate Software & Technologies --------------------------------------------------------------------------- (internet) http://pages.infinit.net/mouser/ mouser@videotron.ca (intranet) http://blackhole.cst.ca/ lavoie@cst.ca --------------------------------------------------------------------------- "No! Try not. Do, or do not. There is no try." -- Yoda on error handling From malcolm at saracen.plsys.co.uk Sun May 16 07:55:47 1999 From: malcolm at saracen.plsys.co.uk (mmalcolm crawford) Date: Thu Nov 3 14:42:03 2005 Subject: DPS && Windows not slow. Message-ID: <199905161455.PAA00643@saracen.plsys.co.uk> David wrote: > > Display PostScript is quite resource intensive and was > > designed for a systme that effeciently messages and moves memory > > between apps. Windows does neither and DPS based apps suffer accordingly. > > Not true. Our validation tests show that in many real-world examples, with > identical YB source code, that drawing on Windows using DPS is at least as > fast, perhaps about 2% faster, as compared with OpenStep/Intel on identical > Intel hardware. > Umm, you haven't negated Bill's points; namely that DPS is resource-intensive. Sure, as many of us have been saying for a while it's fast, but if you have a machine that has insufficient RAM and ends up swapping a lot (especially if you regularly switch between Win32 and YB apps), it crawls. This is a well-known, oft-reported phenomenon, which the new Quartz graphics architecture should alleviate. mmalc. From maury at OAAI.COM Mon May 17 09:44:11 1999 From: maury at OAAI.COM (Maury Markowitz) Date: Thu Nov 3 14:42:03 2005 Subject: non-rectangular clippaths for EPS images do not work on screen Message-ID: <199905171640.MAA14848@OAAI.COM> > If an EPS image is drawn (via NSImage/NSEPSImageRep) after a > non-rectangular clip path has been set, the image on screen clips to > clip path's bounding box... Yes, this is correct, this topic came up recently here when I noticed the same thing in out app. There was no real solution to the problem, the issue appears to exist because the draw method resets the clip as a part of it's setup. We tried to hook out the setup (there's a method for it) but it continues to do it anyway. The solution in our case was to change our draw method to use a TIFF. I make a singleton image for the cache, and in the draw method simply used NSImage's well designed methods to render the EPS into a TIFF, and then composited the TIFF. On print output we simply render the EPS directly. It worked like a champ, I'm very happy with the results and the speed was fine. Maury From estone at knowledgepoint.com Mon May 17 09:05:50 1999 From: estone at knowledgepoint.com (Evan K. Stone) Date: Thu Nov 3 14:42:03 2005 Subject: WWDC Attendance... Message-ID: <000001bea07f$22439de0$ce1615d1@estone.knowledgepoint.com> Any figures on how many attendees there were at WWDC? Unfortunately, I was not able to attend... t h a n k s ! :: e v a n k. s t o n e :: p e t a l u m a, c a l i f o r n i a, u s a ------------------------------------------------ :: mailto:estone@knowledgepoint.com :: http://www.knowledgepoint.com From lavoie at cst.ca Mon May 17 10:34:11 1999 From: lavoie at cst.ca (Martin-Gilles) Date: Thu Nov 3 14:42:03 2005 Subject: WWDC Attendance... Message-ID: <199905171734.NAA18433@plexus.cst.ca> >Any figures on how many attendees there were at WWDC? Unfortunately, I was >not able to attend... > >t h a n k s ! > >:: e v a n k. s t o n e >:: p e t a l u m a, c a l i f o r n i a, u s a >------------------------------------------------ >:: mailto:estone@knowledgepoint.com >:: http://www.knowledgepoint.com 2500 and then some. Up 46% as compared to last year; though, attendence last year was significantly lower than previous years. However, this set a record attendence for any WWDC, according to Apple. =========================================================================== Martin-Gilles Lavoie, Mac OS programmer, Corporate Software & Technologies --------------------------------------------------------------------------- (internet) http://pages.infinit.net/mouser/ mouser@videotron.ca (intranet) http://blackhole.cst.ca/ lavoie@cst.ca --------------------------------------------------------------------------- "No! Try not. Do, or do not. There is no try." -- Yoda on error handling From ckane at apple.com Mon May 17 10:40:06 1999 From: ckane at apple.com (Chris Kane) Date: Thu Nov 3 14:42:03 2005 Subject: CoreFou/Fou example and Re: screwed by Apple Message-ID: <008301bea08c$4d752120$9b0eca11@perspicuity.apple.com> Bill Bumgarner bbum@codefab.com writes: > Apple will move Quartz/Cocoa (Cocoa = (Yellow - DPS) + > Quartz) to the Windows platform. At WWDC, it was discussed > several times that Development and Deployment of WebObjects > will remain a first class citizen (i.e. supported and evolved > with the rest of the platform) on Windows. [...] That is certainly one possible interpretation of what I've seen and heard said. Everyone - not to criticize Bill or any of the others trying to synthesize conclusions from statements made at the conference, but remember that such conclusions are not commitments on the part of Apple. It is sometimes hard to remember months or a year later the what Apple actually said and what somebody else concluded. Yes, what Apple said on this at the conference was vague, unsatisfactory, and occasionally perhaps a bit contradictory. And I can't throw any additional information into the mix, because I don't know any more than what the WWDC attendees know. >Now, combine all of the above with the complete shambles that is the >Windows market. To what API should the quartz engineering be done >against??? The existing Windows NT/98 split APIs-- building the light >weight window management and display infrastructure will require two sets >of significant engineering. [...] This is certainly true, and I'd encourage folks mailing leadership@apple.com or getting together petitions or whatever to be as specific as possible on the range of acceptable or desired options in what you want. For example here, do you need NT/2000 only or do you need 95 and 98 as well? Do you need i386 or would NT on Alpha be sufficient? If Mac OS X ran on i386, would you not need Windows support anymore? (I'm just making these questions up to give examples, not suggesting any future directions. :-)) Actually, I'm not sure why I'm bothering to say this, since the First Law of Software is that customers and developers want everything. (Actually this comes from economics: human needs are finite, human wants and desires are unlimited.) Obligatory -dev content: >- Anyone have example code that shows use of the same core foundation >structure / object from both ANSI C and ObjC? The following two functions are line-for-line equivalent. A line from one function can be replaced with the same line from the other, and it doesn't matter which line or lines you choose. (You will get type warnings from the compiler in some cases.) #import void foundation(void) { NSMutableDictionary *dict = [[NSMutableDictionary allocWithZone:NULL] init]; NSDate *now = [[NSDate allocWithZone:NULL initWithTimeIntervalSinceReferenceDate:[NSDate timeIntervalSinceReferenceDate]]; NSDate *wasCurrent; [dict setObject:now forKey:@"CurrentTime"]; [now release]; /* blah blah blah */ wasCurrent = [dict objectForKey:@"CurrentTime"]; now = [[NSDate allocWithZone:NULL] initWithTimeIntervalSinceReferenceDate:[NSDate timeIntervalSinceReferenceDate]]; if (NSOrderedAscending == [now compare:wasCurrent]) { NSLog(@"Time seems to have gone backwards!"); } else { NSLog(@"Time marches on."); } [now release]; [dict release]; } #include void corefoundation(void) { CFMutableDictionaryRef dict = CFDictionaryCreateMutable(NULL, 0, &kCFTypeDictionaryKeyCallBacks, &kCFTypeDictionaryValueCallBacks); CFDateRef now = CFDateCreate(NULL, CFAbsoluteTimeGetCurrent()); CFDateRef wasCurrent; CFDictionarySetValue(dict, CFSTR("CurrentTime"), now); CFRelease(now); /* blah blah blah */ wasCurrent = CFDictionaryGetValue(dict, CFSTR("CurrentTime")); now = CFDateCreate(NULL, CFAbsoluteTimeGetCurrent()); if (kCFCompareLessThan == CFDateCompare(now, wasCurrent, NULL)) { NSLog(CFSTR("Time seems to have gone backwards!")); } else { NSLog(CFSTR("Time marches on.")); } CFRelease(now); CFRelease(dict); } Notes: * The interchangeability is only possible because of the toll-free bridging. Not all obvious type/class correspondences are toll-free bridged nor will ever be. See release notes in DP1. * Since CFStrings/NSStrings are toll-free bridged, the expressions @"CurrentTime and CFSTR("CurrentTime") are also interchangeable, but the @"" syntax is only valid when an ObjC file is being parsed (i.e., not in a .c file unless the -ObjC switch is given to the compiler). * There is also a CFLog() function, similar to NSLog(), but it is not public yet as we haven't decided how to define one of its parameters, and its implementation is not as complete as NSLog(). * Foundation.h includes CoreFoundation.h, so the #include of the latter is not strictly necessary in the example above. * You get the CoreFoundation framework linked in on Mac OS X with "-framework CoreFoundation" (or adding it to the frameworks list in PB). Since Foundation links with CoreFoundation, apps linked with Foundation implicitly get CoreFoundation brought in. * CoreFoundation is not available in Mac OS X Server 1.1, only in Mac OS X and likely Mac OS X Server releases thereafter. Chris Kane Application Frameworks Apple Computer, Inc. Comments made by drones, such as myself, are not official statements on the part of Apple Computer, Inc. From cdouty at netcom.com Mon May 17 10:47:38 1999 From: cdouty at netcom.com (Chris Douty) Date: Thu Nov 3 14:42:03 2005 Subject: Mac OS 8.6 In-Reply-To: <373CAF3D.5B0A76EA@home.com> Message-ID: On Fri, 14 May 1999, David P. Riedel wrote: > I've been running MacOS 8.6 on my bluebox since b3 of 8.6 -- no > problems. i think the TIL is referring to the disk image that osx sets > up, not an external operating system on a real disk. > > dave riedel > > John Jay Feiler wrote: > > > > From: Randy Wigginton > > > I know Apple recommends against it, but I already installed 8.6 in > > my blue > > > box partition. It seems fine to me -- anybody else out there try > > it yet? > > > --randy Just another data point. I installed 8.6 on the Blue Box within hours of its official release. I haven't had a problem yet, except that my "Help" menu is still titled "Hammerfest" which was the code of my last seed before my ADC membership ran out. Wierd. This is installed on a separate partition and not a disk image. -Chris -- Christopher Douty - Rogue Engineer trapped in a land of software cdouty@netcom.com "Frequently the messages have meaning; that is they refer to or are correlated according to some system with physical or conceptual entities. These semantic aspects of communication are irrelevant to the engineering problem." -Shannon From ajetha at tacticalstep.com Mon May 17 11:00:39 1999 From: ajetha at tacticalstep.com (A Jetha) Date: Thu Nov 3 14:42:03 2005 Subject: WWDC Attendance... Message-ID: <77271BC5159B.AAA5050@bismuth.tor.sfl.net> > Any figures on how many attendees there were at WWDC? Unfortunately, I was > not able to attend... > Something like 2500 developers, 43% better than last year. -- Alykhan Jetha (AJ) TacticalStep Inc., WWW: http://www.tacticalstep.com Email: ajetha@tacticalstep.com Phone: 905.831.2643 ext 1 Fax: 905.837.2546 From maury at OAAI.COM Mon May 17 11:28:44 1999 From: maury at OAAI.COM (Maury Markowitz) Date: Thu Nov 3 14:42:03 2005 Subject: Debugging Nightmares on OSXServer Message-ID: <199905171824.OAA15034@OAAI.COM> > Is anyone having success debugging on OSXServer? Don't get me started... I talked to gdb-man at the show and he's pretty unhappy with the current state of affairs too. Things did improve significantly with the actual Server release, but as you note it's hardly perfect. He knows this, and he's working on it literally night and day. Let me put it this way, he didn't solve any of the specific problems I was having, but I sure felt a lot better about it and I think it's going to get a lot better. In my case my main problem appears to be that the firing up of our licensing thread makes it go wonky. A few pause/conts gets it running, and after that it seems to work pretty well. Are you on the release, or DR2? Maury From Jerry.Porter at targetbase.com Mon May 17 12:36:08 1999 From: Jerry.Porter at targetbase.com (Jerry.Porter@targetbase.com) Date: Thu Nov 3 14:42:03 2005 Subject: WOInfoCenter Message-ID: <00013AC5.C21479@targetbase.com> Hello there, Does anyone know of a way to get WOInfoCenter to read/index/recognize othe documentation or directory trees. I would like to add our documentation and source trees for the ARM project to WOInfoCenter to allow the developers to access the source and docs in a convienient way. I seem to recall someone posted an answer to this a while back. Jerry From bierman at apple.com Mon May 17 12:28:16 1999 From: bierman at apple.com (Peter Bierman) Date: Thu Nov 3 14:42:03 2005 Subject: DP1 & MOSXS on diff disks? In-Reply-To: Message-ID: >Some Dev Questions: > >- Can you stick an EIDE drive in a Beige computer along with the U/W SCSI >drives and boot Mac OS X off of the EIDE, Mac OS X Server off of the U/W >SCSI card? Yes, as long as the IDE disk isn't in slave mode. -pmb -- "UNIX shells in Mac OS X should be unneeded but functional... and have the same installed base as MPW." From kc at omnigroup.com Mon May 17 13:23:24 1999 From: kc at omnigroup.com (Ken Case) Date: Thu Nov 3 14:42:03 2005 Subject: [Moderator] Reminder: this list is for developing software Message-ID: <199905172023.NAA20611@ignem.omnigroup.com> I'd like to remind people that this list is for us to help each other develop software. We have two other lists available for other topics; please use them when appropriate. In particular, both the "screwed by Apples marketing/deceit" and the "when do us non-attendees get the developer preview" threads belong on MacOSX-talk. (The initial question about the developer preview schedule was reasonably appropriate here, but not the "why do we pay $500/year" part.) Please continue those threads elsewhere, not here. I understand that this sort of information is interesting to Mac OS X developers, but so is a lot of other information on a lot of other lists. This list isn't for every topic which interests developers, it's specifically for helping each other write code for Mac OS X. Thank you. Your host, Ken From Philippe.Robert at uptime.ch Mon May 17 13:58:16 1999 From: Philippe.Robert at uptime.ch (Philippe C.D. Robert) Date: Thu Nov 3 14:42:03 2005 Subject: Mac OS X questions... [namely, when do us non-attendees get In-Reply-To: <199905162149.RAA08350@digifix.com> References: <199905162149.RAA08350@digifix.com> Message-ID: <9905172058.AA01129@uptime.ch> You wrote: > Does nobody comprehend that a 300Mb download of a CD image just > isn't viable? > This ain't Mac OS. You need to have a full CD to start the install from. No argument, Darwin is about the same size. sweet dreams, Phil --- Philippe C.D. Robert Uptime Object Factory Inc http://www.nice.ch/~phip Software Development http://www.projectcenter.ch OpenStep/Unix From truong at netmagic.net Mon May 17 14:05:51 1999 From: truong at netmagic.net (Thai Truong) Date: Thu Nov 3 14:42:03 2005 Subject: Mac OS 8.6 In-Reply-To: Message-ID: I seem to be having the same problem. Anyone else experience this? -Thai >Just another data point. I installed 8.6 on the Blue Box within hours of >its official release. I haven't had a problem yet, except that my >"Help" menu is still titled "Hammerfest" which was the code of my last >seed before my ADC membership ran out. Wierd. This is installed on a >separate partition and not a disk image. > > -Chris From sanguish at digifix.com Mon May 17 14:11:23 1999 From: sanguish at digifix.com (Scott Anguish) Date: Thu Nov 3 14:42:03 2005 Subject: Mac OS X questions... [namely, when do us non-attendees get In-Reply-To: <199905162149.RAA08350@digifix.com> Message-ID: <199905172108.RAA14760@digifix.com> > You wrote: > > Does nobody comprehend that a 300Mb download of a CD image just > > isn't viable? > > This ain't Mac OS. You need to have a full CD to start the install from. > > No argument, Darwin is about the same size. Not hardly.. Darwin contains none of the documentation, none of the user apps, none of window server, none of the development applications, None of the AppKit/Foundation/Carbon frameworks. It doesn't contain Java or the other parts either. There is a major difference between the two in size. -- Scott Anguish Middlebury College - Center for Educational Technology Stepwise Server - Mac OS X Server Information From tbumgarner at pobox.com Mon May 17 15:48:48 1999 From: tbumgarner at pobox.com (Tim Bumgarner) Date: Thu Nov 3 14:42:03 2005 Subject: MacOS X info from ADC News Message-ID: <1285169387-250366878@macitools.symantec.com> I just received the ADC news #155 which contained the following info: MAC OS X ---------------------------------------- NOTE: One copy of the Mac OS X Developer Preview 1 CD will be mailed directly to members with an ADC Seed CD asset. Look for these to arrive starting the first week of June. twb From hunter at lastonepicked.com Mon May 17 16:08:42 1999 From: hunter at lastonepicked.com (Hunter Hillegas) Date: Thu Nov 3 14:42:03 2005 Subject: MacOS X info from ADC News Message-ID: <19990517230730611.AAA524@nts1.webhostingprovider.com@firewall.jacobstern.com> Of course. The week my membership expires... Maybe I will slip in... ---------- >From: "Tim Bumgarner" >To: Multiple recipients of list >Subject: MacOS X info from ADC News >Date: Mon, May 17, 1999, 3:58 PM > > I just received the ADC news #155 which contained the following info: > > MAC OS X > ---------------------------------------- > NOTE: One copy of the Mac OS X Developer Preview 1 CD will be mailed > directly to members with an ADC Seed CD asset. Look for these to > arrive starting the first week of June. > > > > twb > From wilkie at cg.tuwien.ac.at Mon May 17 16:29:54 1999 From: wilkie at cg.tuwien.ac.at (Alexander Wilkie) Date: Thu Nov 3 14:42:03 2005 Subject: teTeX - anyone? Message-ID: <199905172329.BAA01761@vilya.cg.tuwien.ac.at> Hi! (I searched the list archives for previous postings on this, but no avail.) My problem: I am trying to get teTeX 0.4pl7 (a public domain LaTeX package) to compile under OSXS. While this is not strictly speaking MacOSX development, but (from an Apple perspective) hacking with unclean command-line tools, I still ask this forum for advice. The build almost works, but not quite, and maybe someone can help me out with the missing bit. After manual twiddling with the makefiles, I manage to get the package to compile, except for the interactive configure utility (which is unfortunately pretty essential), which uses ncurses. Unfortunately, it uses a version of ncurses (1.99e, the source comes with the teTeX source tree) which is not quite the same thing as that which comes with OSXS (the manpage says that it is ncurses under the name curses) - some procedure definitions do not match. The situation is that ncurses 1.99e that comes with teTeX does not compile, and the OSX ncurses can´t be used without hacking the dialog utility (apparently, at least). I even got ncurses 1.99g to compile cleanly (in order to replace the abridged 1.99e that comes with teTeX with a whole library), but the result of this effort does not work - it reports "errors opening the terminal d0" - and trying to fix this with a broken OSXS debugger and without knowing much about terminal libraries to begin with is not my idea of how to spend time in a meaningful way. My only incentive to get started with this in the first place was to get a working, minimalist TeX on my G3 - locally, for personal usage, no clean port for the outside world. No big effort. (Hah!) I do not assume that this can´t be made to work - I´m certainly no expert w/r to terminal libaries, and probably I´ve overlooked something obvious. But better to expose oneself as clueless on a mailinglist and save a few days more of trying, than to pretend to be able to do it on one´s own. Hrm. Any input is welcome. ys Alexander Wilkie -- e-mail : mailto:wilkie@cg.tuwien.ac.at www : http://www.cg.tuwien.ac.at/staff/AlexanderWilkie.html pgp : finger wilkie@www.cg.tuwien.ac.at From nemesys at ix.netcom.com Mon May 17 21:05:56 1999 From: nemesys at ix.netcom.com (Laurent Daudelin) Date: Thu Nov 3 14:42:03 2005 Subject: Implementing Registration Serial Number References: <9905171446.AA18395@vviuh221.vvi.com> Message-ID: <3740E723.83CE5CB2@ix.netcom.com> Hi all! I'm looking for some suggestion/example source code on how to implement a registration serial number in my application. I don't really need a floating license thing, but I'd be willing to have a look if someone has any suggestion. Thanks in advance! -Laurent. -- ============================================================ Laurent Daudelin http://www.nemesys-software.com Logiciels Nemesys Software mailto:nemesys@ix.netcom.com -------------- next part -------------- A non-text attachment was scrubbed... Name: nemesys.vcf Type: text/x-vcard Size: 323 bytes Desc: Card for Laurent Daudelin Url : /mailman/archive/macosx-dev/attachments/19990518/5e5896cc/nemesys.vcf From sammis at its.caltech.edu Mon May 17 22:46:47 1999 From: sammis at its.caltech.edu (Ian Sammis) Date: Thu Nov 3 14:42:03 2005 Subject: teTeX - anyone? Message-ID: <199905180546.WAA07068@chamber.cco.caltech.edu> > From: Alexander Wilkie > Date: 1999-05-17 16:43:27 -0700 > To: Multiple recipients of list > Subject: teTeX - anyone? > X-Mailer: by Apple MailViewer (2.106) > X-Listprocessor-Version: 6.0 -- ListProcessor by Anastasios Kotsikonas > Originator: macosx-dev@omnigroup.com > X-Comment: To unsubscribe, follow directions at > http://www.omnigroup.com/MailArchive/ > Howdy, While I've never tried teTeX, I was able to get the web2c sources to compile quite nicely under X Server, so if your goal isn't so much to get teTeX as to get a nice solid latex engine going you might try that... From fan at chem.ucla.edu Mon May 17 23:31:00 1999 From: fan at chem.ucla.edu (Paul S. Fan) Date: Thu Nov 3 14:42:03 2005 Subject: teTeX - anyone? Message-ID: <199905180631.XAA03365@krypton.chem.ucla.edu> Hi - > I am trying to get teTeX .. to compile Bill Chin has compiled this. I have mine installed along with the Sun, SGI and HP binaries. I believe there is a link to this at www.stepwise.com. I am still hoping for a port of a TexViewer from someone soon :-) Best, Paul From sanguish at digifix.com Mon May 17 23:52:09 1999 From: sanguish at digifix.com (Scott Anguish) Date: Thu Nov 3 14:42:03 2005 Subject: Additional WWDC Coverage notes now on Stepwise Message-ID: <199905180649.CAA16990@digifix.com> I've got all the remainder of my notes, and some additional stuff forwarded by Tomi Engel, Maury Markowitz and Mmalc. Tonight I added content about the Mac OS X file systems, networking, app layout, iokit and generally cleaned up my notes. If anyone has notes about the NetInfo sessions that they'd like to submit, or about the WOF sessions.. I'd love to get a copy... There is much technical stuff there, so I hope that this isn't out of bounds in the dev list.. Oh, I also finished uploading the rest of the images( 20 or so) to the Rogues Gallery.. -- Scott Anguish Middlebury College - Center for Educational Technology Stepwise Server - Mac OS X Server Information From uli at www.kaben-net.de Tue May 18 00:18:06 1999 From: uli at www.kaben-net.de (Ulrich Koester) Date: Thu Nov 3 14:42:03 2005 Subject: Mac OS 8.6 Message-ID: <199905180718.JAA15121@www.kaben-net.de> Hi The Til article is right. With 8.5 installed a crashed BlueBox can be quit via Command-Shift-Q in 99.99%. With 8.6 installed a crashed BlueBox crashes the whole system in about 30%. The MacOS X Sever is realy down. No connection via ftp. telnet. http and so on. Command-Shift-Q doesent work. The is no response on hitting Control-PowerOn. The only thing left is reboot. This is on a beige G3. Good luck Ulrich -- Ulrich Köster Heino Haase Segel GmbH Gneversdorfer Weg 9 23570 Travemünde Germany fon: +49 4502 2038 fax: +49 4502 2315 From malte at oops.se Tue May 18 02:31:40 1999 From: malte at oops.se (Malte Tancred) Date: Thu Nov 3 14:42:03 2005 Subject: SOLVED: Subprojects in PB legacy projects Message-ID: I asked for a way to make PB aware of subprojects in an arbitrary c source tree, ie when creating a Legacy project. I asked before trying anything out on my own. It is as simple as adding a subproject the normal way in PB, then simply specifying the Legacy subproject type. I'll be a bit more careful the next time... *blush* Cheerio, Malte -- Malte Tancred, OOPS ab malte@oops.se http://www.oops.se From clandolt at datacomm.ch Tue May 18 02:59:12 1999 From: clandolt at datacomm.ch (Ch. Landolt) Date: Thu Nov 3 14:42:03 2005 Subject: looking for converter programs Message-ID: <199905180959.LAA13332@smtp.datacomm.ch> hi, i am currently developing a web based application that converts files into different formats. for that purpose i am looking for unix programs running under macosxs that convert files. specifically, i am looking for a ps2pdf converter, a mp3 converter and image converters. as i am no unix geek, i am not really into building those apps myself, so i would be gratefull, if you could point me to binaries.... cheers, chris From theisen at akaMail.com Tue May 18 03:10:43 1999 From: theisen at akaMail.com (Dirk Theisen) Date: Thu Nov 3 14:42:03 2005 Subject: teTeX - anyone? In-Reply-To: <199905180631.XAA03365@krypton.chem.ucla.edu> Message-ID: <1drzvnm.xqhu3912jrrheM@ascend-tk-p8.rhrz.uni-bonn.de> Hi! > > I am trying to get teTeX .. to compile > > Bill Chin has compiled this. I have mine installed along > with the Sun, SGI and HP binaries. I believe there is a > link to this at www.stepwise.com. > > I am still hoping for a port of a TexViewer from someone > soon :-) I was hoping someone would also compile the pdfTeX extension, so we'd had a viewer (e.g. OmniPDF) already. With the upcoming PDF integration this seems like a good idea anyway. Samuel, as you have already compiled the sources: Could you include the pdfTeX extension and distribute a binary? This would be /really/ nice to have!! Please do it! Regards, Dirk Some Links: http://www.tug.org/applications/pdftex/ ftp://ftp.cstug.cz/pub/tex/local/cstug/thanh/pdftex-beta/latest/ A version that can include (some forms of) Tiffs which should be most interesting for X Server: ftp://ftp.fi.muni.cz/pub/tex/local/cstug/janik/pdftex-0.13d-tiff www.kiffe.com/tex/pdftex.html But it should also be available at CTAN. -- Buy a Pentium III now, get your personal Big Brother FREE!!! http://theisen.home.pages.de/ From lavoie at cst.ca Tue May 18 05:29:54 1999 From: lavoie at cst.ca (Martin-Gilles) Date: Thu Nov 3 14:42:03 2005 Subject: Additional WWDC Coverage notes now on Stepwise Message-ID: <199905181229.IAA21421@plexus.cst.ca> >I've got all the remainder of my notes, and some additional stuff >forwarded by Tomi Engel, Maury Markowitz and Mmalc. I'd like to thank you, Scott and the others, on behalf of those that couldn't attent. =========================================================================== Martin-Gilles Lavoie, Mac OS programmer, Corporate Software & Technologies --------------------------------------------------------------------------- (internet) http://pages.infinit.net/mouser/ mouser@videotron.ca (intranet) http://blackhole.cst.ca/ lavoie@cst.ca --------------------------------------------------------------------------- "No! Try not. Do, or do not. There is no try." -- Yoda on error handling From mvgfr at netcom.com Tue May 18 05:50:11 1999 From: mvgfr at netcom.com (Marc Farnum Rendino) Date: Thu Nov 3 14:42:03 2005 Subject: Mac OS X questions... [namely, when do us non-attendees get the In-Reply-To: References: Message-ID: At 9:24 PM -0700 on 99/05/10, John Ray wrote: >Has Apple indicated if/when those of us who aren't attending WWDC will see the >preview release 1 of Mac OS X? I certainly hope it's included in the May or >June seedings... I just saw a notice that said it'll be coming in early June, to those with a seed key. - Marc From foobar at objectfarm.org Tue May 18 06:51:07 1999 From: foobar at objectfarm.org (Stefan Kreutter) Date: Thu Nov 3 14:42:03 2005 Subject: Lombard and MacOS X Server ? Message-ID: Hi, does anyone know if MacOS X Server runs on a Lombard PowerBook G3 ? I'd like to buy one but only if MacOS X Server does install and work. -Stefan From cpry at enc.org Tue May 18 07:21:33 1999 From: cpry at enc.org (chad pry) Date: Thu Nov 3 14:42:03 2005 Subject: configuration Message-ID: <000701bea139$bae992b0$7b619280@enc.org> is there a way to lock the screen while logged into an account on rhapsody. also. is there a way to add a user from the terminal. i remember nextstep had a command line step to do this but i havent been able to find it with rhapsody. chad From ajetha at tacticalstep.com Tue May 18 07:33:11 1999 From: ajetha at tacticalstep.com (A Jetha) Date: Thu Nov 3 14:42:03 2005 Subject: Additional WWDC Coverage notes now on Stepwise Message-ID: <77271C26AC0.AAA2FD5@bismuth.tor.sfl.net> > >I've got all the remainder of my notes, and some additional stuff > >forwarded by Tomi Engel, Maury Markowitz and Mmalc. > > I'd like to thank you, Scott and the others, on behalf of those that > couldn't attent. > In that vein, we should be thanking Scott (and the contributor & sponsors) for all the work that has gone into maintaining and updating StepWise over the years! Thanks Scott!!! -- Alykhan Jetha (AJ) TacticalStep Inc., WWW: http://www.tacticalstep.com Email: ajetha@tacticalstep.com Phone: 905.831.2643 ext 1 Fax: 905.837.2546 From epeyton at epicware.com Tue May 18 08:54:35 1999 From: epeyton at epicware.com (Eric S. Peyton) Date: Thu Nov 3 14:42:03 2005 Subject: read and write on Windows Message-ID: All, I am currently porting a unix based system to YB/windows. This system uses a lot of file descriptor code using write() and read(), etc. Does anyone know the easiest way to port this stuff? I have no familiarity with winsock or windows NT really and don't even know where to start. All the other code in the app ported easily just using winsock but these calls don't seem to have a straight replacement. Eric Eric Peyton epeyton@epicware.com Software and Source for Mac OS X Server From portnoy at ai.mit.edu Tue May 18 09:10:53 1999 From: portnoy at ai.mit.edu (Stephen Peters) Date: Thu Nov 3 14:42:03 2005 Subject: teTeX - anyone? In-Reply-To: theisen@akaMail.com's message of "Tue, 18 May 1999 03:27:42 -0700 (PDT)" References: <1drzvnm.xqhu3912jrrheM@ascend-tk-p8.rhrz.uni-bonn.de> Message-ID: theisen@akaMail.com (Dirk Theisen) writes: > > > I am trying to get teTeX .. to compile > > > > Bill Chin has compiled this. I have mine installed along > > with the Sun, SGI and HP binaries. I believe there is a > > link to this at www.stepwise.com. > > > > I am still hoping for a port of a TexViewer from someone > > soon :-) > > I was hoping someone would also compile the pdfTeX extension, so we'd > had a viewer (e.g. OmniPDF) already. > > With the upcoming PDF integration this seems like a good idea > anyway. Actually, I read somewhere that pdftex is going to be part of the next major teTeX release (0.9, currently in beta). Maybe I should download the beta teTeX and start playing with it... For my current viewing needs, I've been inserting a little bit of code into the dvips header so that it turns on antialiasing command to make the output look really nice under Preview.app. Unfortunately, `setantialias' doesn't handle the bitmap'd fonts as nicely as it does Times-Roman or other PostScript fonts, but it's an improvement. -- Stephen L. Peters portnoy@ai.mit.edu PGP fingerprint: BFA4 D0CF 8925 08AE 0CA5 CCDD 343D 6AC6 "Poodle: The other white meat." -- Sherman, Sherman's Lagoon From cpry at enc.org Tue May 18 09:53:14 1999 From: cpry at enc.org (chad pry) Date: Thu Nov 3 14:42:03 2005 Subject: strange? Message-ID: <000901bea14e$eb769210$7b619280@enc.org> after i make/compile a application, is there a special way to copy the .app directory to another location? i just tried copying FreeSpace.app from the /tmp directory, i could not do it with workspace manager. so i tried with a command line and the application will not run. chad From epeyton at epicware.com Tue May 18 10:27:30 1999 From: epeyton at epicware.com (Eric S. Peyton) Date: Thu Nov 3 14:42:03 2005 Subject: Read and write on Windows Message-ID: All, The solution that worked for me was to replace the write() and read() calls with similiar send() and recv() calls in winsock. The parameters to these calls only differ in the last flags argument, which is defined in WinSock.h. Thanks to everyone for their help, Eric "Eric S. Peyton" wrote: > > All, > > I am currently porting a unix based system to YB/windows. This system > uses a > lot of file descriptor code using write() and read(), etc. Does anyone > know the easiest way to port this stuff? I have no familiarity with > winsock or windows NT really and don't even know where to start. All the > other code in the app ported easily just using winsock but these calls > don't seem to have a straight replacement. > > Eric > > Eric Peyton > epeyton@epicware.com Eric Peyton epeyton@epicware.com Software and Source for Mac OS X Server From lavoie at cst.ca Tue May 18 10:48:53 1999 From: lavoie at cst.ca (Martin-Gilles) Date: Thu Nov 3 14:42:03 2005 Subject: strange? Message-ID: <199905181748.NAA23117@plexus.cst.ca> >after i make/compile a application, is there a special way to copy the .app >directory to another location? i just tried copying FreeSpace.app from the >/tmp directory, i could not do it with workspace manager. so i tried with a >command line and the application will not run. > >chad You may not have the rights/permissions to move your app to the target directory. Try logging in as root. =========================================================================== Martin-Gilles Lavoie, Mac OS programmer, Corporate Software & Technologies --------------------------------------------------------------------------- (internet) http://pages.infinit.net/mouser/ mouser@videotron.ca (intranet) http://blackhole.cst.ca/ lavoie@cst.ca --------------------------------------------------------------------------- "No! Try not. Do, or do not. There is no try." -- Yoda on error handling From maury at OAAI.COM Tue May 18 12:01:45 1999 From: maury at OAAI.COM (Maury Markowitz) Date: Thu Nov 3 14:42:03 2005 Subject: JPEG codex Message-ID: <199905181858.OAA16815@OAAI.COM> > Use 'TIFFRepresentationUsingCompression:factor:' from NSImage and specify > JPEG compression. I just finished adding GIF support - five lines plus some debugging, gotta love it - and I'm curious about adding more document types in future revs. What I'm a bit confused by is the API. Does the call above produce what is basically a JFIF file? If I use that call, then take the data and put it to disk, can I rename it to .jfif and have it work? Or would the above call produce a TIFF that's compressed with JPEG and still needs a TIFF viewer to open? Also has anyone had any luck with some of the other formats (BMP, PNG etc.)? GIF "just worked", but the rest don't seem to be as easy to use. Maury From aa4lr at radio.org Tue May 18 14:37:50 1999 From: aa4lr at radio.org (Bill Coleman AA4LR) Date: Thu Nov 3 14:42:03 2005 Subject: screwed by Apples marketing/deceit Message-ID: <990418173750.RAA00980@gate.iterated.com> On 5/16/99 19:27, Jorge Mederos at jmederos@gran-via.com wrote: >At least in the enterprise market this is true. The only light on the >enterprise horizont is WebObjects. Why? because it's from Apple? No. >Because it's good and it's cross-platform. It's not very difficult to >get people receptive to WebObjects because it's deployed on Sun or HP >servers, developed (at least until now) on NT or MXS, supports all >important DB, CORBA and can be written in Java or Obj-C. If WebObjects >moves, or if there is a risk or fear that WebObjects may move towards a >more propietary Apple hardware, much of its good mindshare will be lost. I think the WWDC may have sent a message to Apple. Most of the WO sessions were standing room only. The fact that it garnered such intense interest can't be lost on Apple. >Bad thing is that. If Apple wish to have any possibility in the >enterprise market in few years, they must begin now. They must begin >repairing broken paths. If they only care about selling iMacs to >secretaries... ok, they only need to care about Microsoft supporting >Office on MacOS X. If there is not Office, there wont be iMacs in the >enterprise either. Well, it can work the other way around. At lunch one day at WWDC, I happened to sit with an engineer from the Microsoft Mac Business Unit. He happened to let slip that Microsoft was planning to soon release a version of Outlook (not Express) for the Mac. This new version with have virtual feature parity with the Windows 9x version. (For those who haven't had the (mis)fortune of using Outlook, it scarcely has one third of the features on the Mac as it does in the Windows version) As a user of Outlook on the Mac, this was good news. It was also interesting how it came about. Seems that some of Microsoft's Outlook users had a smattering of Macs. There's a conflict between the way Outlook on Windows 9x manipulates the calendar and the way Outlook on Mac does it. In short, they don't interoperate very well. These users handed Microsoft an ultimatum: make Outlook on Mac work correctly, or we'll pull it from our enterprise. To Microsoft's credit, they listened. Talk about the tail wagging the dog.... >What path? One that continues past strategies. If Apple doesn't want to >sell YB-NT (oops! Cocoa) for the masses yet... well, then there should >be an alternative. >From the feedback session with the VPs, it appeared that Apple may have gotten this message, but haven't yet figured out how they want to go about it. Consider that MacOS X Server didn't have a price set until the very last possible moment. Bill Coleman, AA4LR, PP-ASEL Mail: aa4lr@radio.org Quote: "Not within a thousand years will man ever fly!" -- Wilbur Wright, 1901 From rcfa at cubiculum.com Tue May 18 14:06:52 1999 From: rcfa at cubiculum.com (Ronald C.F. Antony) Date: Thu Nov 3 14:42:03 2005 Subject: strange? In-Reply-To: <000901bea14e$eb769210$7b619280@enc.org> References: <000901bea14e$eb769210$7b619280@enc.org> Message-ID: <199905182106.RAA15484@kannix.cubiculum.com> you wrote: > after i make/compile a application, is there a special way to copy the .app > directory to another location? i just tried copying FreeSpace.app from the > /tmp directory, i could not do it with workspace manager. so i tried with a > command line and the application will not run. The easiest way is to set the installation location in PB, and then do a make install. For copying, you need the proper permissions, and if copying with the CLI, you need to specify the proper flags to do a recursive copy if you copy an app wrapper, otherwise the wrapper's directory may be copied as a flat file, which is not what you want. Ronald ============================================================================== "The reasonable man adapts himself to the world; the unreasonable one persists in trying to adapt the world to himself. Therefore all progress depends on the unreasonable man." G.B. Shaw | rcfa@cubiculum.com | NeXT-mail welcome From philippe at point-net.com Tue May 18 15:32:38 1999 From: philippe at point-net.com (Philippe Duval) Date: Thu Nov 3 14:42:03 2005 Subject: ftpchroot Message-ID: Anyone have try to create the '/private/etc/ftpchroot' file to limit access of ftp server? It does not have any effect on my ftpd preinstall on MacOSX. I put in it the login users names. thanks ___________________________________ Philippe Duval Point Net Communication http://www.point-net.com/ From cwhite at comnetix.com Tue May 18 16:43:28 1999 From: cwhite at comnetix.com (Craig White) Date: Thu Nov 3 14:42:03 2005 Subject: JPEG codex In-Reply-To: <199905181858.OAA16815@OAAI.COM> Message-ID: <9905181943.AA284003@deli> No, unfortunately NSImage just gives you a TIFF wrapper with JPEG compression inside of it. This is part of the TIFF specs but it is not the same file format as a JFIF. JFIF is considered the 'Open' jpeg file standard. Check out the Independent JPEG Group for more info and source code. Craig From pf at ipagents.com Wed May 19 01:16:11 1999 From: pf at ipagents.com (Peter Fagerlund) Date: Thu Nov 3 14:42:03 2005 Subject: Existing EOF Adaptors on OSXS Message-ID: What EOF Adaptors for MacOSXS are available today ? ... and where is it documented / tech spec ? or am I to develop on another system to get the full range ? ... Thanks /peter From gclem at frontline-software.dk Wed May 19 01:30:58 1999 From: gclem at frontline-software.dk (Geert B. Clemmensen) Date: Thu Nov 3 14:42:03 2005 Subject: Existing EOF Adaptors on OSXS References: Message-ID: <199905190831.KAA01101@snaps.frontline-software.dk> Peter Fagerlund wrote: What EOF Adaptors for MacOSXS are available today ? ... and where is it documented / tech spec ? or am I to develop on another system to get the full range ? ... You can get EOF adaptors for FrontBase (www.frontline-software.com) and OpenBase (www.openbase.com). You are welcome to download a copy of FrontBase v. 1.1a, it is still free a couple more days. FrontBase 1.2 will be available early next week, but there will still be a free version. Geert B. Clemmensen Frontline Software www.frontline-software.com -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: text/enriched Size: 563 bytes Desc: not available Url : /mailman/archive/macosx-dev/attachments/19990519/4b6ade90/attachment.bin From sbrandon at music.gla.ac.uk Wed May 19 03:53:25 1999 From: sbrandon at music.gla.ac.uk (Stephen Brandon - SysAdmin) Date: Thu Nov 3 14:42:03 2005 Subject: Sending mouse events programatically Message-ID: <199905191053.LAA03833@music.gla.ac.uk> Hi, I'm using OpenStep4.2 at the moment, but I guess this is the same on Rhapsody etc. I programatically create a region in a view on screen, which when clicked on creates a field editor and starts editing. What I want to happen is for the editing to start automatically, without the user having to physically click on it. No matter what I do I can't seem to get NSApp or the window object to send a programatically created event to my view in the window. I have tried [NSApp postEvent:atStart:], and [NSApp sendEvent:] (which is cautioned against in the docs). The location of the event I create is definitely correct. When creating mouse click events, how does one generate a valid event number? If I take the current event number and add 1 to it, will that upset the next event to be created by the system? Could this be the problem? Since the action of creating the new region is started by the user clicking on a button on another window, could the DPSContext be wrong? I'm getting it with [NSApp context]. I have tried sending the mouseDown event to the [view mouseDown:] method via performSelector:afterDelay:, which performs the click ok, but I can't seem to get the mouseUp event to register. After receiving the mouseDown, the application hangs (spinning cursor), as I guess it is waiting for the mouseUp. If I click on the titlebar of the window, editing resumes normally. I have tried sending the mouseUp event to the window, to the view, and to the field editor (all delayed so they get there .02s after the mouseDown), but none have any effect. Help! Thanks, Stephen Brandon SysAdmin, Music Dept, Glasgow University From sbrandon at music.gla.ac.uk Wed May 19 04:14:37 1999 From: sbrandon at music.gla.ac.uk (Stephen Brandon - SysAdmin) Date: Thu Nov 3 14:42:03 2005 Subject: RTF data and NSTextView Message-ID: <199905191114.MAA03900@music.gla.ac.uk> Hi, 2nd horrible problem of the day: In the mould of Draw.app, I edit text in a field editor on screen, which is then stored internally as NSData of the RTF pulled from the field editor after editing finishes. When drawing the view on screen, the NSData RTF is stuck into a resident NSTextView and rendered on screen. The problem is that the when rendering rtf back on screen, words frequently wrap where they didn't in the field editor. Specifically, 1-letter words on the end of the line often get wrapped to the next line. This would seem to indicate that either (a) the text is drawn with different attributes (eg kerning), or (b) the bounds, frame or text container size are different between the field editor and the resident NSTextView I use for rendering. I am sure that (a) does not apply, and I have checked all sizes, bounds, lineFragmentPadding etc and I'm sure they are identical. To work around the problem, I currently add 2 to the view width after editing has finished, but this then messes up any right-aligned text. And every time you edit it, the view grows some more. Note: the field editor auto-expands, but the displaying NSTextView just takes the final size of the field editor as a fixed size horizontally, and sticks the text into it. Could this be causing the difference in behaviour? Can anyone shed any light on why this is happening, and what to do about it? Is rtf the best way to store the data (as in Draw.app), or should I just keep a copy of the NSTextContainer/NSAttributedString? Many thanks in advance for any help, Stephen Brandon From gerard at ina.fr Wed May 19 04:37:03 1999 From: gerard at ina.fr (Gerard Iglesias) Date: Thu Nov 3 14:42:03 2005 Subject: JPEG codex References: <9905181943.AA284003@deli> Message-ID: <3742A25E.DBB7AB15@ina.fr> Craig White wrote: > > No, unfortunately NSImage just gives you a TIFF wrapper with JPEG compression inside of it. This is part of the TIFF specs but it is not the same file format as a JFIF. JFIF is considered the 'Open' jpeg file standard. Check out the Independent JPEG Group for more info and source code. > > Craig I was trying to use Jpeg compression with NSImage on WO4/NT, but I got an exception about broken link with the image server. Any help? Sincerely. -- Gerard Iglesias Ph.D. in Computer Graphic gerard.iglesias@ina.fr Researcher/Project Manager http://www.mediaport.net/People/Gerard.IGLESIAS From Philippe.Robert at uptime.ch Wed May 19 05:02:15 1999 From: Philippe.Robert at uptime.ch (Philippe Robert) Date: Thu Nov 3 14:42:03 2005 Subject: MOSXS and UFS Message-ID: <9905191202.AA20339@california.uptime.ch> Hi! I have the following problem: I set up my B&W G3 with 4 partitions. I then installed MOSXS on the first partition. Now the 3 others seem to be HFS+ (I partitioned it with MacOS 8.5...*grrrr*), and I can't initialise those partitions with UFS anymore?! Any hints? ... thx in advance! sweet dreams, Phil -- Philippe C.D. Robert | Uptime ObjectFactory Inc Unix/OpenStep Software Engineer | http://www.uptime.ch http://www.nice.ch/~phip | info@uptime.ch From sbrandon at music.gla.ac.uk Wed May 19 05:25:11 1999 From: sbrandon at music.gla.ac.uk (Stephen Brandon - SysAdmin) Date: Thu Nov 3 14:42:03 2005 Subject: RTF data and NSTextView In-Reply-To: <199905191114.MAA03900@music.gla.ac.uk> References: <199905191114.MAA03900@music.gla.ac.uk> Message-ID: <199905191225.NAA04160@music.gla.ac.uk> OK, found a good way to transfer rtf from one NSTextView to another without it going and reformatting itself. In the receiving NSTextView, don't set the NSTextContainer to track the NSTextView size horizontally. Instead, manually set it to the size of the NSTextView where necessary, + a bit (using 2 at the moment, will try to be more accurate). The lineFragmentPadding (default 5) keeps the text within the bounds. Hope this helps someone else; sorry for responding to my own post. Stephen Brandon From andrew at stone.com Wed May 19 09:18:19 1999 From: andrew at stone.com (Andrew Stone) Date: Thu Nov 3 14:42:04 2005 Subject: RTF data and NSTextView Message-ID: <199905191626.KAA26970@floyd.stone.com> Don't use RTF - use the much richer NSAttributedString. While you are at, don't use "Draw" - use Mike Ferris's "Sketch" as a much cleaner example. ;-) Andrew === Andrew Stone, QuixMaster of Stone Design Corp mailto:andrew@stone.com (505) 345-4800 http://www.stone.com - Create(TM) From andrew at stone.com Wed May 19 09:22:45 1999 From: andrew at stone.com (Andrew Stone) Date: Thu Nov 3 14:42:04 2005 Subject: AppleScripting from OSXS? Message-ID: <199905191631.KAA26985@floyd.stone.com> At the Cocoa talks on Friday last week at WWDC, someone mentioned that they had haqued a way to send AppleEvents from the Yellow Box. This would be more ideal than testing scriptability from the Blue Box - since you'd be able to watch as the script executed. Whoever that was, would you mind updating us on what you have and it's availability? Thanks! === Andrew Stone, QuixMaster of Stone Design Corp mailto:andrew@stone.com (505) 345-4800 http://www.stone.com - Create(TM) From mtrent at msn.fullfeed.com Wed May 19 09:38:35 1999 From: mtrent at msn.fullfeed.com (Mike Trent) Date: Thu Nov 3 14:42:04 2005 Subject: MOSXS and UFS Message-ID: <199905191638.LAA23331@fullfeed.msn.fullfeed.com> > I have the following problem: I set up my B&W G3 with 4 partitions. I then > installed MOSXS on the first partition. Now the 3 others seem to be HFS+ (I > partitioned it with MacOS 8.5...*grrrr*), and I can't initialise those > partitions with UFS anymore?! Any hints? ... > > thx in advance! Due to the way the system works, you may have one-and-only-one Apple_Rhapsody_UFS partition. The kernel is going to look for one Apple_Rhapsody_UFS partition (either the first or the last, I don't know which), look for a traditional UNIX disklabel, and use that for partition information. I suppose it might be possible to hack up your disklabel to point to these other Apple Partition Map entries, but I haven't tried it. You would need to write a program that lets you read and write your disklabel and you would need to spend some time in pdisk. Mike Trent From Philippe.Robert at uptime.ch Wed May 19 09:50:59 1999 From: Philippe.Robert at uptime.ch (Philippe Robert) Date: Thu Nov 3 14:42:04 2005 Subject: MOSXS and UFS In-Reply-To: <199905191638.LAA23331@fullfeed.msn.fullfeed.com> References: <199905191638.LAA23331@fullfeed.msn.fullfeed.com> Message-ID: <9905191650.AA21534@california.uptime.ch> You wrote: [snip] > I suppose it might be possible to hack up your disklabel to point to these > other Apple Partition Map entries, but I haven't tried it. You would need > to write a program that lets you read and write your disklabel and you > would need to spend some time in pdisk. Yep, I know this from OPENSTEP, but I thought I heard of a 'hack' regarding this issue..?! I could be wrong, though. Thx anyway! sweet dreams, Phil -- Philippe C.D. Robert | Uptime ObjectFactory Inc Unix/OpenStep Software Engineer | http://www.uptime.ch http://www.nice.ch/~phip | info@uptime.ch From maury at OAAI.COM Wed May 19 10:06:29 1999 From: maury at OAAI.COM (Maury Markowitz) Date: Thu Nov 3 14:42:04 2005 Subject: testing software for YB Message-ID: <199905191702.NAA18573@OAAI.COM> At WWDC I mentioned I had heard of a testing system for OS/YB/Cocoa apps. Here's the link, I hope this gets back to the person that was asking... http://www.sente.ch/software/ocunit/ Maury From wchin at acm.org Wed May 19 10:31:08 1999 From: wchin at acm.org (Bill Chin) Date: Thu Nov 3 14:42:04 2005 Subject: Samba dbench and MOXS profiling Message-ID: <199905191634.MAA06139@mail.richmond.net> I did a port of Andrew Tridgell's dbench for testing disk throughput with multiple clients for Samba. (see ftp://samba.org/pub/tridge/dbench/README). MOXS's performance was abysmal, which is in direct contradiction to the results from another disk benchmark, Bonnie. I need some help analyzing the profiler output. I have a 400Mhz Blue and Ice G3, 256mb RAM, Apple OEM Symbios Logic 875 U/W SCSI controller, 9.1gb IBM Ultrastar 9LZX 10k rpm 4mb cache U2 drive. Under MacOS with the ATTO benchmark, the system gets about 25mb/sec burst throughput and 16-18mb/sec sustained throughput. Running bonnie under MOXS, which tries to be more clever about cache busting than iozone, I get similar sustained numbers using a 1 or 2 gigabyte test file (to help reduce the effect of memory caching). My first port of dbench just replaces the simple SYSV shared memory implementation with a Mach vm_inherit setup. Quite easy, actually. dbench forks off a number of clients (which can be spec'ed), uses the shared memory system to sync them, then lets them loose executing a script makes each client mimic the operations done with the benchmark NetBench. MOXS's performance on my setup with a single client was about 3mb/sec. On my 500Mhz 1mb L3 cache Alpha PC164 machine running Debian Linux w/ 64mb RAM, kernel 2.2.6, same SCSI controller, 4.5gb IBM Ultrastar 2ES 5400 rpm, 512k cache, Ultra SCSI, gets _22_mb/sec. Obviously the Alpha is getting caching results, since the disk can only do about 5-7mb sustained throughput. Basically, on a machine that is supposed to have 50% less integer and disk performance, is beating the crap out of the G3. So I compiled the benchmark with profiling and ran into my first problem. Since dbench forks a bunch of clients, I only get the gmon.out for the main process, not the interesting clients. How does one get a separate profiler output for each child? Well, to get around that and remove suspicions around my shared memory hack, I modified the benchmark to only run one client w/o forking. Performance is the same. The gprof output is below. Questions and observations: 1) Why isn't the %time for _main 100%? It is under Linux 2) A good 40% is spent on interpreting the script to run the benchmark, a little bit higher than on Linux on the Alpha. 3) I expect a lot more time in the actual file operation functions. That, apparently didn't happen or I'm doing something wrong here. Thanks! call graph profile: The sum of self and descendents is the major sort for this listing. function entries: index the index of the function in the call graph listing, as an aid to locating it (see below). %time the percentage of the total time of the program accounted for by this function and its descendents. self the number of seconds spent in this function itself. descendents the number of seconds spent in the descendents of this function on behalf of this function. called the number of times this function is called (other than recursive calls). self the number of times this function calls itself recursively. name the name of the function, with an indication of its membership in a cycle, if any. index the index of the function in the call graph listing, as an aid to locating it. parent listings: self* the number of seconds of this function's self time which is due to calls from this parent. descendents* the number of seconds of this function's descendent time which is due to calls from this parent. called** the number of times this function is called by this parent. This is the numerator of the fraction which divides up the function's time to its parents. total* the number of times this function was called by all of its parents. This is the denominator of the propagation fraction. parents the name of this parent, with an indication of the parent's membership in a cycle, if any. index the index of this parent in the call graph listing, as an aid in locating it. children listings: self* the number of seconds of this child's self time which is due to being called by this function. descendent* the number of seconds of this child's descendent's time which is due to being called by this function. called** the number of times this child is called by this function. This is the numerator of the propagation fraction for this child. total* the number of times this child is called by all functions. This is the denominator of the propagation fraction. children the name of this child, and an indication of its membership in a cycle, if any. index the index of this child in the call graph listing, as an aid to locating it. * these fields are omitted for parents (or children) in the same cycle as the function. If the function (or child) is a member of a cycle, the propagated times and propagation denominator represent the self time and descendent time of the cycle as a whole. ** static-only parents and children are indicated by a call count of 0. cycle listings: the cycle as a whole is listed with the same fields as a function entry. Below it are listed the members of the cycle, and their contributions to the time and call counts of the cycle. granularity: each sample hit covers 4 byte(s) for 0.63% of 1.58 seconds called/total parents index %time self descendents called+self name index called/total children 0.00 1.17 1/1 __start [2] [1] 74.1 0.00 1.17 1 _main [1] 0.00 1.17 1/1 _do_one_child [3] 0.00 0.00 1/101216 _atoi [10] 0.00 0.00 1/42 _printf [37] ----------------------------------------------- [2] 74.1 0.00 1.17 __start [2] 0.00 1.17 1/1 _main [1] ----------------------------------------------- 0.00 1.17 1/1 _main [1] [3] 74.1 0.00 1.17 1 _do_one_child [3] 0.18 0.99 1/1 _child_run [4] 0.00 0.00 1/1 _start_timer [46] 0.00 0.00 1/1 _end_timer [45] 0.00 0.00 3/42 _printf [37] 0.00 0.00 1/1 _setbuffer [282] ----------------------------------------------- 0.18 0.99 1/1 _do_one_child [3] [4] 74.1 0.18 0.99 1 _child_run [4] 0.05 0.35 289605/289605 _all_string_sub [5] 0.25 0.00 610173/610173 _strtok [7] 0.01 0.12 96536/96536 _fgets [9] 0.01 0.09 101215/101216 _atoi [10] 0.00 0.04 6393/6393 _do_stat [20] 0.00 0.02 3624/3626 _do_open [22] 0.01 0.01 23287/23287 _do_write [23] 0.01 0.00 13765/13765 _strcpy [26] 0.00 0.01 961/961 _do_unlink [29] 0.00 0.00 195/195 _do_rename [30] 0.00 0.00 13763/697394 _strncmp [12] 0.00 0.00 15/15 _do_mkdir [31] 0.00 0.00 10/10 _do_rmdir [32] 0.00 0.00 2/2 _do_create [38] 0.00 0.00 2/2 _sprintf [47] 0.00 0.00 1/42 _printf [37] 0.00 0.00 4698/4698 _do_read [240] 0.00 0.00 3617/3619 _do_close [241] 0.00 0.00 1/1 _signal [287] 0.00 0.00 1/1 _fopen [278] 0.00 0.00 1/1 _fclose [276] ----------------------------------------------- 0.05 0.35 289605/289605 _child_run [4] [5] 25.6 0.05 0.35 289605 _all_string_sub [5] 0.21 0.09 394046/394046 _strstr [6] 0.03 0.00 104441/104441 _memmove [21] 0.03 0.00 104441/202011 _memcpy [16] ----------------------------------------------- 0.21 0.09 394046/394046 _all_string_sub [5] [6] 18.9 0.21 0.09 394046 _strstr [6] 0.09 0.00 683609/697394 _strncmp [12] ----------------------------------------------- 0.25 0.00 610173/610173 _child_run [4] [7] 15.8 0.25 0.00 610173 _strtok [7] ----------------------------------------------- [8] 11.4 0.18 0.00 _strlen [8] ----------------------------------------------- 0.01 0.12 96536/96536 _child_run [4] [9] 8.5 0.01 0.12 96536 _fgets [9] 0.10 0.00 97511/97514 _memchr [11] 0.02 0.00 97511/202011 _memcpy [16] 0.00 0.00 1007/1007 ___srefill [243] ----------------------------------------------- 0.00 0.00 1/101216 _main [1] 0.01 0.09 101215/101216 _child_run [4] [10] 6.3 0.01 0.09 101216 _atoi [10] 0.09 0.00 101216/101216 _strtol [13] ----------------------------------------------- 0.00 0.00 3/97514 ___sfvwrite [35] 0.10 0.00 97511/97514 _fgets [9] [11] 6.3 0.10 0.00 97514 _memchr [11] ----------------------------------------------- 0.00 0.00 22/697394 ___findenv [40] 0.00 0.00 13763/697394 _child_run [4] 0.09 0.00 683609/697394 _strstr [6] [12] 5.7 0.09 0.00 697394 _strncmp [12] ----------------------------------------------- 0.09 0.00 101216/101216 _atoi [10] [13] 5.7 0.09 0.00 101216 _strtol [13] ----------------------------------------------- [14] 5.7 0.09 0.00 _lseek [14] ----------------------------------------------- 0.00 0.00 10/11395 _do_rmdir [32] 0.00 0.00 15/11395 _do_mkdir [31] 0.00 0.00 390/11395 _do_rename [30] 0.00 0.00 961/11395 _do_unlink [29] 0.01 0.01 3626/11395 _do_open [22] 0.02 0.02 6393/11395 _do_stat [20] [15] 4.4 0.03 0.04 11395 _strupper [15] 0.04 0.00 410233/410233 _toupper [18] ----------------------------------------------- 0.00 0.00 59/202011 ___sfvwrite [35] 0.02 0.00 97511/202011 _fgets [9] 0.03 0.00 104441/202011 _all_string_sub [5] [16] 3.2 0.05 0.00 202011 _memcpy [16] ----------------------------------------------- [17] 3.2 0.05 0.00 _write [17] ----------------------------------------------- 0.04 0.00 410233/410233 _strupper [15] [18] 2.5 0.04 0.00 410233 _toupper [18] ----------------------------------------------- [19] 2.5 0.04 0.00 _strcmp [19] ----------------------------------------------- 0.00 0.04 6393/6393 _child_run [4] [20] 2.5 0.00 0.04 6393 _do_stat [20] 0.02 0.02 6393/11395 _strupper [15] ----------------------------------------------- 0.03 0.00 104441/104441 _all_string_sub [5] [21] 1.9 0.03 0.00 104441 _memmove [21] ----------------------------------------------- 0.00 0.00 2/3626 _do_create [38] 0.00 0.02 3624/3626 _child_run [4] [22] 1.4 0.00 0.02 3626 _do_open [22] 0.01 0.01 3626/11395 _strupper [15] 0.00 0.00 37/42 _printf [37] ----------------------------------------------- 0.01 0.01 23287/23287 _child_run [4] [23] 1.3 0.01 0.01 23287 _do_write [23] 0.01 0.00 51/51 _memset [27] ----------------------------------------------- [24] 1.3 0.02 0.00 _close [24] ----------------------------------------------- [25] 1.3 0.02 0.00 _fstat [25] ----------------------------------------------- 0.01 0.00 13765/13765 _child_run [4] [26] 0.6 0.01 0.00 13765 _strcpy [26] ----------------------------------------------- 0.01 0.00 51/51 _do_write [23] [27] 0.6 0.01 0.00 51 _memset [27] ----------------------------------------------- [28] 0.6 0.01 0.00 _read [28] ----------------------------------------------- 0.00 0.01 961/961 _child_run [4] [29] 0.4 0.00 0.01 961 _do_unlink [29] 0.00 0.00 961/11395 _strupper [15] ----------------------------------------------- 0.00 0.00 195/195 _child_run [4] [30] 0.2 0.00 0.00 195 _do_rename [30] 0.00 0.00 390/11395 _strupper [15] ----------------------------------------------- 0.00 0.00 15/15 _child_run [4] [31] 0.0 0.00 0.00 15 _do_mkdir [31] 0.00 0.00 15/11395 _strupper [15] ----------------------------------------------- 0.00 0.00 10/10 _child_run [4] [32] 0.0 0.00 0.00 10 _do_rmdir [32] 0.00 0.00 10/11395 _strupper [15] ----------------------------------------------- [33] 0.0 0.00 0.00 44+80 [33] 0.00 0.00 84 _vfprintf [34] 0.00 0.00 40 ___sbprintf [248] ----------------------------------------------- 40 ___sbprintf [248] 0.00 0.00 2/44 _sprintf [47] 0.00 0.00 42/44 _printf [37] [34] 0.0 0.00 0.00 84 _vfprintf [34] 0.00 0.00 48/48 ___sprint [36] 0.00 0.00 3/3 _isinf [259] 0.00 0.00 3/3 _isnan [260] 0.00 0.00 3/3 _cvt [257] 0.00 0.00 3/3 ___ultoa [256] 0.00 0.00 1/1 ___swsetup [272] 40 ___sbprintf [248] ----------------------------------------------- 0.00 0.00 48/48 ___sprint [36] [35] 0.0 0.00 0.00 48 ___sfvwrite [35] 0.00 0.00 59/202011 _memcpy [16] 0.00 0.00 3/97514 _memchr [11] 0.00 0.00 2/42 _fflush [247] ----------------------------------------------- 0.00 0.00 48/48 _vfprintf [34] [36] 0.0 0.00 0.00 48 ___sprint [36] 0.00 0.00 48/48 ___sfvwrite [35] ----------------------------------------------- 0.00 0.00 1/42 _main [1] 0.00 0.00 1/42 _child_run [4] 0.00 0.00 3/42 _do_one_child [3] 0.00 0.00 37/42 _do_open [22] [37] 0.0 0.00 0.00 42 _printf [37] 0.00 0.00 42/44 _vfprintf [34] ----------------------------------------------- 0.00 0.00 2/2 _child_run [4] [38] 0.0 0.00 0.00 2 _do_create [38] 0.00 0.00 2/3626 _do_open [22] 0.00 0.00 2/3619 _do_close [241] ----------------------------------------------- 0.00 0.00 1/2 _start_timer [46] 0.00 0.00 1/2 _end_timer [45] [39] 0.0 0.00 0.00 2 _gettimeofday [39] 0.00 0.00 1/1 _localtime [43] ----------------------------------------------- 0.00 0.00 1/1 _getenv [41] [40] 0.0 0.00 0.00 1 ___findenv [40] 0.00 0.00 22/697394 _strncmp [12] 0.00 0.00 1/1 __NSGetEnviron [267] ----------------------------------------------- 0.00 0.00 1/1 _tzset [44] [41] 0.0 0.00 0.00 1 _getenv [41] 0.00 0.00 1/1 ___findenv [40] ----------------------------------------------- 0.00 0.00 1/1 _localtime [43] [42] 0.0 0.00 0.00 1 _localsub [42] 0.00 0.00 1/1 _tzset [44] 0.00 0.00 1/1 _timesub [290] ----------------------------------------------- 0.00 0.00 1/1 _gettimeofday [39] [43] 0.0 0.00 0.00 1 _localtime [43] 0.00 0.00 1/1 _localsub [42] ----------------------------------------------- 0.00 0.00 1/1 _localsub [42] [44] 0.0 0.00 0.00 1 _tzset [44] 0.00 0.00 1/1 _getenv [41] 0.00 0.00 1/1 _tzsetwall [292] ----------------------------------------------- 0.00 0.00 1/1 _do_one_child [3] [45] 0.0 0.00 0.00 1 _end_timer [45] 0.00 0.00 1/2 _gettimeofday [39] ----------------------------------------------- 0.00 0.00 1/1 _do_one_child [3] [46] 0.0 0.00 0.00 1 _start_timer [46] 0.00 0.00 1/2 _gettimeofday [39] ----------------------------------------------- 0.00 0.00 2/2 _child_run [4] [47] 0.0 0.00 0.00 2 _sprintf [47] 0.00 0.00 2/44 _vfprintf [34] ----------------------------------------------- 0.00 0.00 4698/4698 _child_run [4] [240] 0.0 0.00 0.00 4698 _do_read [240] ----------------------------------------------- 0.00 0.00 2/3619 _do_create [38] 0.00 0.00 3617/3619 _child_run [4] [241] 0.0 0.00 0.00 3619 _do_close [241] ----------------------------------------------- 0.00 0.00 1007/1007 ___srefill [243] [242] 0.0 0.00 0.00 1007 ___sread [242] ----------------------------------------------- 0.00 0.00 1007/1007 _fgets [9] [243] 0.0 0.00 0.00 1007 ___srefill [243] 0.00 0.00 1007/1007 ___sread [242] 0.00 0.00 1/2 ___smakebuf [262] ----------------------------------------------- 0.00 0.00 242/242 _tzload [291] [244] 0.0 0.00 0.00 242 _detzcode [244] ----------------------------------------------- 0.00 0.00 1/43 _setvbuf [284] 0.00 0.00 42/43 _fflush [247] [245] 0.0 0.00 0.00 43 ___sflush [245] 0.00 0.00 42/42 ___swrite [246] ----------------------------------------------- 0.00 0.00 42/42 ___sflush [245] [246] 0.0 0.00 0.00 42 ___swrite [246] ----------------------------------------------- 0.00 0.00 2/42 ___sfvwrite [35] 0.00 0.00 40/42 ___sbprintf [248] [247] 0.0 0.00 0.00 42 _fflush [247] 0.00 0.00 42/43 ___sflush [245] ----------------------------------------------- 40 _vfprintf [34] [248] 0.0 0.00 0.00 40 ___sbprintf [248] 0.00 0.00 40/42 _fflush [247] 40 _vfprintf [34] ----------------------------------------------- 0.00 0.00 11/11 cerror [234] [249] 0.0 0.00 0.00 11 _cthread_set_errno_self [249] ----------------------------------------------- 0.00 0.00 3/6 _d2b [258] 0.00 0.00 3/6 ___dtoa [255] [250] 0.0 0.00 0.00 6 _Balloc [250] 0.00 0.00 2/4 _malloc [253] ----------------------------------------------- 0.00 0.00 5/5 ___dtoa [255] [251] 0.0 0.00 0.00 5 _Bfree [251] ----------------------------------------------- 0.00 0.00 1/5 _nxzonefreenolock [265] 0.00 0.00 4/5 _nxzonemallocnolock [254] [252] 0.0 0.00 0.00 5 _downheap [252] ----------------------------------------------- 0.00 0.00 2/4 ___smakebuf [262] 0.00 0.00 2/4 _Balloc [250] [253] 0.0 0.00 0.00 4 _malloc [253] 0.00 0.00 4/4 _nxzonemallocnolock [254] ----------------------------------------------- 0.00 0.00 4/4 _malloc [253] [254] 0.0 0.00 0.00 4 _nxzonemallocnolock [254] 0.00 0.00 4/5 _downheap [252] 0.00 0.00 2/2 _nxzonefreenolock [265] ----------------------------------------------- 0.00 0.00 3/3 _cvt [257] [255] 0.0 0.00 0.00 3 ___dtoa [255] 0.00 0.00 5/5 _Bfree [251] 0.00 0.00 3/3 _d2b [258] 0.00 0.00 3/6 _Balloc [250] ----------------------------------------------- 0.00 0.00 3/3 _vfprintf [34] [256] 0.0 0.00 0.00 3 ___ultoa [256] ----------------------------------------------- 0.00 0.00 3/3 _vfprintf [34] [257] 0.0 0.00 0.00 3 _cvt [257] 0.00 0.00 3/3 ___dtoa [255] ----------------------------------------------- 0.00 0.00 3/3 ___dtoa [255] [258] 0.0 0.00 0.00 3 _d2b [258] 0.00 0.00 3/6 _Balloc [250] 0.00 0.00 3/3 _lo0bits [261] ----------------------------------------------- 0.00 0.00 3/3 _vfprintf [34] [259] 0.0 0.00 0.00 3 _isinf [259] ----------------------------------------------- 0.00 0.00 3/3 _vfprintf [34] [260] 0.0 0.00 0.00 3 _isnan [260] ----------------------------------------------- 0.00 0.00 3/3 _d2b [258] [261] 0.0 0.00 0.00 3 _lo0bits [261] ----------------------------------------------- 0.00 0.00 1/2 ___srefill [243] 0.00 0.00 1/2 ___swsetup [272] [262] 0.0 0.00 0.00 2 ___smakebuf [262] 0.00 0.00 2/2 ___swhatbuf [263] 0.00 0.00 2/4 _malloc [253] 0.00 0.00 1/1 _isatty [279] ----------------------------------------------- 0.00 0.00 2/2 ___smakebuf [262] [263] 0.0 0.00 0.00 2 ___swhatbuf [263] ----------------------------------------------- 0.00 0.00 1/2 _fclose [276] 0.00 0.00 1/2 _setvbuf [284] [264] 0.0 0.00 0.00 2 _free [264] ----------------------------------------------- 0.00 0.00 2/2 _nxzonemallocnolock [254] [265] 0.0 0.00 0.00 2 _nxzonefreenolock [265] 0.00 0.00 2/2 _upheap [266] 0.00 0.00 1/1 _findregionnolock [277] 0.00 0.00 1/5 _downheap [252] 0.00 0.00 1/1 _vm_deallocate [293] ----------------------------------------------- 0.00 0.00 2/2 _nxzonefreenolock [265] [266] 0.0 0.00 0.00 2 _upheap [266] ----------------------------------------------- 0.00 0.00 1/1 ___findenv [40] [267] 0.0 0.00 0.00 1 __NSGetEnviron [267] 0.00 0.00 1/1 __dyld_lookup_and_bind [274] ----------------------------------------------- 0.00 0.00 1/1 _fclose [276] [268] 0.0 0.00 0.00 1 ___sclose [268] ----------------------------------------------- 0.00 0.00 1/1 _fopen [278] [269] 0.0 0.00 0.00 1 ___sflags [269] ----------------------------------------------- 0.00 0.00 1/1 _fopen [278] [270] 0.0 0.00 0.00 1 ___sfp [270] ----------------------------------------------- 0.00 0.00 1/1 ___swsetup [272] [271] 0.0 0.00 0.00 1 ___sinit [271] ----------------------------------------------- 0.00 0.00 1/1 _vfprintf [34] [272] 0.0 0.00 0.00 1 ___swsetup [272] 0.00 0.00 1/1 ___sinit [271] 0.00 0.00 1/2 ___smakebuf [262] ----------------------------------------------- 0.00 0.00 1/1 _sigaction__ [286] [273] 0.0 0.00 0.00 1 __dyld_bind_fully_image_containing_address [273] ----------------------------------------------- 0.00 0.00 1/1 __NSGetEnviron [267] [274] 0.0 0.00 0.00 1 __dyld_lookup_and_bind [274] ----------------------------------------------- 0.00 0.00 1/1 __mcleanup [838] [275] 0.0 0.00 0.00 1 _exit [275] ----------------------------------------------- 0.00 0.00 1/1 _child_run [4] [276] 0.0 0.00 0.00 1 _fclose [276] 0.00 0.00 1/1 ___sclose [268] 0.00 0.00 1/2 _free [264] ----------------------------------------------- 0.00 0.00 1/1 _nxzonefreenolock [265] [277] 0.0 0.00 0.00 1 _findregionnolock [277] ----------------------------------------------- 0.00 0.00 1/1 _child_run [4] [278] 0.0 0.00 0.00 1 _fopen [278] 0.00 0.00 1/1 ___sflags [269] 0.00 0.00 1/1 ___sfp [270] ----------------------------------------------- 0.00 0.00 1/1 ___smakebuf [262] [279] 0.0 0.00 0.00 1 _isatty [279] 0.00 0.00 1/1 _tcgetattr [289] ----------------------------------------------- 0.00 0.00 1/1 _vm_deallocate [293] [280] 0.0 0.00 0.00 1 _mig_get_reply_port [280] ----------------------------------------------- 0.00 0.00 1/1 _vm_deallocate [293] [281] 0.0 0.00 0.00 1 _msg_rpc [281] ----------------------------------------------- 0.00 0.00 1/1 _do_one_child [3] [282] 0.0 0.00 0.00 1 _setbuffer [282] 0.00 0.00 1/1 _setvbuf [284] ----------------------------------------------- 0.00 0.00 1/1 _tzsetwall [292] [283] 0.0 0.00 0.00 1 _settzname [283] ----------------------------------------------- 0.00 0.00 1/1 _setbuffer [282] [284] 0.0 0.00 0.00 1 _setvbuf [284] 0.00 0.00 1/43 ___sflush [245] 0.00 0.00 1/2 _free [264] ----------------------------------------------- 0.00 0.00 1/1 _signal__ [288] [285] 0.0 0.00 0.00 1 _sigaction [285] 0.00 0.00 1/1 _sigaction__ [286] ----------------------------------------------- 0.00 0.00 1/1 _sigaction [285] [286] 0.0 0.00 0.00 1 _sigaction__ [286] 0.00 0.00 1/1 __dyld_bind_fully_image_containing_address [273] ----------------------------------------------- 0.00 0.00 1/1 _child_run [4] [287] 0.0 0.00 0.00 1 _signal [287] 0.00 0.00 1/1 _signal__ [288] ----------------------------------------------- 0.00 0.00 1/1 _signal [287] [288] 0.0 0.00 0.00 1 _signal__ [288] 0.00 0.00 1/1 _sigaction [285] ----------------------------------------------- 0.00 0.00 1/1 _isatty [279] [289] 0.0 0.00 0.00 1 _tcgetattr [289] ----------------------------------------------- 0.00 0.00 1/1 _localsub [42] [290] 0.0 0.00 0.00 1 _timesub [290] ----------------------------------------------- 0.00 0.00 1/1 _tzsetwall [292] [291] 0.0 0.00 0.00 1 _tzload [291] 0.00 0.00 242/242 _detzcode [244] ----------------------------------------------- 0.00 0.00 1/1 _tzset [44] [292] 0.0 0.00 0.00 1 _tzsetwall [292] 0.00 0.00 1/1 _tzload [291] 0.00 0.00 1/1 _settzname [283] ----------------------------------------------- 0.00 0.00 1/1 _nxzonefreenolock [265] [293] 0.0 0.00 0.00 1 _vm_deallocate [293] 0.00 0.00 1/1 _mig_get_reply_port [280] 0.00 0.00 1/1 _msg_rpc [281] ----------------------------------------------- flat profile: % the percentage of the total running time of the time program used by this function. cumulative a running sum of the number of seconds accounted seconds for by this function and those listed above it. self the number of seconds accounted for by this seconds function alone. This is the major sort for this listing. calls the number of times this function was invoked, if this function is profiled, else blank. self the average number of milliseconds spent in this ms/call function per call, if this function is profiled, else blank. total the average number of milliseconds spent in this ms/call function and its descendents per call, if this function is profiled, else blank. name the name of the function. This is the minor sort for this listing. The index shows the location of the function in the gprof listing. If the index is in parenthesis it shows where it would appear in the gprof listing if it were to be printed. granularity: each sample hit covers 4 byte(s) for 0.43% of 2.35 seconds % cumulative self self total time seconds seconds calls ms/call ms/call name 22.6 0.53 0.53 _moncount (2236) 10.6 0.78 0.25 610173 0.00 0.00 _strtok [7] 10.2 1.02 0.24 mcount (237) 8.9 1.23 0.21 394046 0.00 0.00 _strstr [6] 7.7 1.41 0.18 1 180.00 1169.99 _child_run [4] 7.7 1.59 0.18 _strlen [8] 4.3 1.69 0.10 97514 0.00 0.00 _memchr [11] 3.8 1.78 0.09 697394 0.00 0.00 _strncmp [12] 3.8 1.87 0.09 101216 0.00 0.00 _strtol [13] 3.8 1.96 0.09 _lseek [14] 2.1 2.01 0.05 289605 0.00 0.00 _all_string_sub [5] 2.1 2.06 0.05 202011 0.00 0.00 _memcpy [16] 2.1 2.11 0.05 _write [17] 1.7 2.15 0.04 410233 0.00 0.00 _toupper [18] 1.7 2.19 0.04 _strcmp [19] 1.3 2.22 0.03 104441 0.00 0.00 _memmove [21] 1.3 2.25 0.03 11395 0.00 0.01 _strupper [15] 0.9 2.27 0.02 _close [24] 0.9 2.29 0.02 _fstat [25] 0.4 2.30 0.01 101216 0.00 0.00 _atoi [10] 0.4 2.31 0.01 96536 0.00 0.00 _fgets [9] 0.4 2.32 0.01 23287 0.00 0.00 _do_write [23] 0.4 2.33 0.01 13765 0.00 0.00 _strcpy [26] 0.4 2.34 0.01 51 0.20 0.20 _memset [27] 0.4 2.35 0.01 _read [28] 0.0 2.35 0.00 6393 0.00 0.01 _do_stat [20] 0.0 2.35 0.00 4698 0.00 0.00 _do_read [240] 0.0 2.35 0.00 3626 0.00 0.01 _do_open [22] 0.0 2.35 0.00 3619 0.00 0.00 _do_close [241] 0.0 2.35 0.00 1007 0.00 0.00 ___sread [242] 0.0 2.35 0.00 1007 0.00 0.00 ___srefill [243] 0.0 2.35 0.00 961 0.00 0.01 _do_unlink [29] 0.0 2.35 0.00 242 0.00 0.00 _detzcode [244] 0.0 2.35 0.00 195 0.00 0.01 _do_rename [30] 0.0 2.35 0.00 84 0.00 0.00 _vfprintf [34] 0.0 2.35 0.00 48 0.00 0.00 ___sfvwrite [35] 0.0 2.35 0.00 48 0.00 0.00 ___sprint [36] 0.0 2.35 0.00 43 0.00 0.00 ___sflush [245] 0.0 2.35 0.00 42 0.00 0.00 ___swrite [246] 0.0 2.35 0.00 42 0.00 0.00 _fflush [247] 0.0 2.35 0.00 42 0.00 0.00 _printf [37] 0.0 2.35 0.00 40 0.00 0.00 ___sbprintf [248] 0.0 2.35 0.00 15 0.00 0.01 _do_mkdir [31] 0.0 2.35 0.00 11 0.00 0.00 _cthread_set_errno_self [249] 0.0 2.35 0.00 10 0.00 0.01 _do_rmdir [32] 0.0 2.35 0.00 6 0.00 0.00 _Balloc [250] 0.0 2.35 0.00 5 0.00 0.00 _Bfree [251] 0.0 2.35 0.00 5 0.00 0.00 _downheap [252] 0.0 2.35 0.00 4 0.00 0.00 _malloc [253] 0.0 2.35 0.00 4 0.00 0.00 _nxzonemallocnolock [254] 0.0 2.35 0.00 3 0.00 0.00 ___dtoa [255] 0.0 2.35 0.00 3 0.00 0.00 ___ultoa [256] 0.0 2.35 0.00 3 0.00 0.00 _cvt [257] 0.0 2.35 0.00 3 0.00 0.00 _d2b [258] 0.0 2.35 0.00 3 0.00 0.00 _isinf [259] 0.0 2.35 0.00 3 0.00 0.00 _isnan [260] 0.0 2.35 0.00 3 0.00 0.00 _lo0bits [261] 0.0 2.35 0.00 2 0.00 0.00 ___smakebuf [262] 0.0 2.35 0.00 2 0.00 0.00 ___swhatbuf [263] 0.0 2.35 0.00 2 0.00 0.01 _do_create [38] 0.0 2.35 0.00 2 0.00 0.00 _free [264] 0.0 2.35 0.00 2 0.00 0.00 _gettimeofday [39] 0.0 2.35 0.00 2 0.00 0.00 _nxzonefreenolock [265] 0.0 2.35 0.00 2 0.00 0.00 _sprintf [47] 0.0 2.35 0.00 2 0.00 0.00 _upheap [266] 0.0 2.35 0.00 1 0.00 0.00 __NSGetEnviron [267] 0.0 2.35 0.00 1 0.00 0.00 ___findenv [40] 0.0 2.35 0.00 1 0.00 0.00 ___sclose [268] 0.0 2.35 0.00 1 0.00 0.00 ___sflags [269] 0.0 2.35 0.00 1 0.00 0.00 ___sfp [270] 0.0 2.35 0.00 1 0.00 0.00 ___sinit [271] 0.0 2.35 0.00 1 0.00 0.00 ___swsetup [272] 0.0 2.35 0.00 1 0.00 0.00 __dyld_bind_fully_image_containing_address [273] 0.0 2.35 0.00 1 0.00 0.00 __dyld_lookup_and_bind [274] 0.0 2.35 0.00 1 0.00 1170.00 _do_one_child [3] 0.0 2.35 0.00 1 0.00 0.00 _end_timer [45] 0.0 2.35 0.00 1 0.00 0.00 _exit [275] 0.0 2.35 0.00 1 0.00 0.00 _fclose [276] 0.0 2.35 0.00 1 0.00 0.00 _findregionnolock [277] 0.0 2.35 0.00 1 0.00 0.00 _fopen [278] 0.0 2.35 0.00 1 0.00 0.00 _getenv [41] 0.0 2.35 0.00 1 0.00 0.00 _isatty [279] 0.0 2.35 0.00 1 0.00 0.00 _localsub [42] 0.0 2.35 0.00 1 0.00 0.00 _localtime [43] 0.0 2.35 0.00 1 0.00 1170.00 _main [1] 0.0 2.35 0.00 1 0.00 0.00 _mig_get_reply_port [280] 0.0 2.35 0.00 1 0.00 0.00 _msg_rpc [281] 0.0 2.35 0.00 1 0.00 0.00 _setbuffer [282] 0.0 2.35 0.00 1 0.00 0.00 _settzname [283] 0.0 2.35 0.00 1 0.00 0.00 _setvbuf [284] 0.0 2.35 0.00 1 0.00 0.00 _sigaction [285] 0.0 2.35 0.00 1 0.00 0.00 _sigaction__ [286] 0.0 2.35 0.00 1 0.00 0.00 _signal [287] 0.0 2.35 0.00 1 0.00 0.00 _signal__ [288] 0.0 2.35 0.00 1 0.00 0.00 _start_timer [46] 0.0 2.35 0.00 1 0.00 0.00 _tcgetattr [289] 0.0 2.35 0.00 1 0.00 0.00 _timesub [290] 0.0 2.35 0.00 1 0.00 0.00 _tzload [291] 0.0 2.35 0.00 1 0.00 0.00 _tzset [44] 0.0 2.35 0.00 1 0.00 0.00 _tzsetwall [292] 0.0 2.35 0.00 1 0.00 0.00 _vm_deallocate [293] Index by function name [250] _Balloc [22] _do_open [281] _msg_rpc [251] _Bfree [240] _do_read [265] _nxzonefreenolock [267] __NSGetEnviron [30] _do_rename [254] _nxzonemallocnolock [255] ___dtoa [32] _do_rmdir [37] _printf [40] ___findenv [20] _do_stat [28] _read [248] ___sbprintf [29] _do_unlink [282] _setbuffer [268] ___sclose [23] _do_write [283] _settzname [269] ___sflags [252] _downheap [284] _setvbuf [245] ___sflush [45] _end_timer [285] _sigaction [270] ___sfp [275] _exit [286] _sigaction__ [35] ___sfvwrite [276] _fclose [287] _signal [271] ___sinit [247] _fflush [288] _signal__ [262] ___smakebuf [9] _fgets [47] _sprintf [36] ___sprint [277] _findregionnolock [46] _start_timer [242] ___sread [278] _fopen [19] _strcmp [243] ___srefill [264] _free [26] _strcpy [263] ___swhatbuf [25] _fstat [8] _strlen [246] ___swrite [41] _getenv [12] _strncmp [272] ___swsetup [39] _gettimeofday [6] _strstr [256] ___ultoa [279] _isatty [7] _strtok [273] __dyld_bind_fully_i [259] _isinf [13] _strtol [274] __dyld_lookup_and_b [260] _isnan [15] _strupper [5] _all_string_sub [261] _lo0bits [289] _tcgetattr [10] _atoi [42] _localsub [290] _timesub [4] _child_run [43] _localtime [18] _toupper [24] _close [14] _lseek [291] _tzload [249] _cthread_set_errno_ [1] _main [44] _tzset [257] _cvt [253] _malloc [292] _tzsetwall [258] _d2b [11] _memchr [266] _upheap [244] _detzcode [16] _memcpy [34] _vfprintf [241] _do_close [21] _memmove [293] _vm_deallocate [38] _do_create [27] _memset [17] _write [31] _do_mkdir [280] _mig_get_reply_port (237) mcount [3] _do_one_child (2236) _moncount [33] -- Bill Chin (NeXTmail/MIME accepted) -------------- next part -------------- Skipped content of type multipart/mixed From cwhite at comnetix.com Wed May 19 10:56:24 1999 From: cwhite at comnetix.com (Craig White) Date: Thu Nov 3 14:42:04 2005 Subject: JPEG codex In-Reply-To: <3742A25E.DBB7AB15@ina.fr> Message-ID: <9905191356.AA240071@deli> We used to have a similar problem in NextStep 3.3 days, where a corrupted image would kill the image server. We don't use Jpeg Tiffs anymore, so I'm not sure if the problem still exists, or if it's what you're witnessing. Craig From jkoppi at diehlgraphsoft.com Wed May 19 13:16:07 1999 From: jkoppi at diehlgraphsoft.com (Jeff Koppi) Date: Thu Nov 3 14:42:04 2005 Subject: MacOS X Carbon Examples Message-ID: <1284977807-18445909@diehlgraphsoft.com> Ok I've got "MacOS X Developer Preview 1" installed, updated my CW Pro 4 to 4.1 and installed the tools from the "CW for Mac OS X & Carbon" CD and built the Example applications. The Carbon apps run on my MacOS 8.6 system, but do not run on MacOS X. If I double-click the app I get the following alert: "MuscleDemo is a Mac OS application. To use it, open Mac OS.app then open the application using the Mac OS Finder." If I use a terminal to key in the command "/usr/Carbon/bin/LaunchCFMApp MuscleCarbon" then I get the following error code -42006. (By the way, the mach-0 version of BasicApp did run on MacOS X, but only after I set its execute privileges.) Does anyone have the examples from MW or the Sample Code from Apple running as Carbon apps on MacOS X? Thanks, -- Jeffrey Koppi jkoppi@diehlgraphsoft.com Diehl Graphsoft, Inc. - VectorWorks - award winning CAD for Mac/Win. From sbender at harmony-ds.com Wed May 19 13:16:17 1999 From: sbender at harmony-ds.com (Scott Bender) Date: Thu Nov 3 14:42:04 2005 Subject: ANNOUNCE: MacOSXAmp 0.1 is available Message-ID: <37431C11.CC7A990C@harmony-ds.com> MacOSXAmp is ready for its first public release. MacOSXAmp is a clone of x11amp which is a clone of WinAmp. It looks and feels almost exactly as WinAmp running on Windows, and it supports the exact same skins. The main reason for this release is to get feedback and to, hopefully, get help from others with development. This is a very early version. Although it's close, it is not quite usable. The main problem is the the MPEG decoders are not fast enough. They need optimized for the PPC platform. I would appreciate any help with this and the other things in the TODO and BUGS files. For info, news, binary and source downloads, go to http://www.harmony-ds.com/MacOSXAmp - Scott From tom at evatac.com Wed May 19 13:27:01 1999 From: tom at evatac.com (Tom Carstensen) Date: Thu Nov 3 14:42:04 2005 Subject: MacOS X Carbon Examples Message-ID: <1284976871-4671177@evatac.com> I'd be interested in hearing about this too, as i have apps that run under Carbon lib on Mac OS 8.6, but i do not yet have preview 1 os x. Jeff -- might i ask you to try 2 carbon apps we posted yesterday: ftp://ftp.evatac.com/pub/evatac/beta/SimpleEdit-3.4.3b6(C)-Installer.hqx ftp://ftp.evatac.com/pub/evatac/beta/Snoop-1.6b1(C)-Installer.hqx see if they have the same results. They are Installer Vise installer so you will probably have to copy them over after you install them under 8.x. Thanks -Tom >Ok I've got "MacOS X Developer Preview 1" installed, updated my CW Pro 4 to >4.1 and installed the tools from the "CW for Mac OS X & Carbon" CD and built >the Example applications. > >The Carbon apps run on my MacOS 8.6 system, but do not run on MacOS X. > >If I double-click the app I get the following alert: "MuscleDemo is a Mac OS >application. To use it, open Mac OS.app then open the application using the >Mac OS Finder." > >If I use a terminal to key in the command "/usr/Carbon/bin/LaunchCFMApp >MuscleCarbon" then I get the following error code -42006. > > >(By the way, the mach-0 version of BasicApp did run on MacOS X, but only >after I set its execute privileges.) > >Does anyone have the examples from MW or the Sample Code from Apple running >as Carbon apps on MacOS X? From mhenders at metrowerks.com Wed May 19 13:46:04 1999 From: mhenders at metrowerks.com (Matt Henderson) Date: Thu Nov 3 14:42:04 2005 Subject: MacOS X Carbon Examples Message-ID: <199905192048.PAA10425@Metrowerks.com> >Does anyone have the examples from MW or the Sample Code from Apple running >as Carbon apps on MacOS X? Yes, we do. Read the file "Mac OS X Install CD:Carbon SDK:***README about Carbon***"--specifically the paragraph that says: "In order to launch CFM based executables from an HFS drive you will need to split the resource and data forks using the "UFS Converter" tool which is included with the Carbon development kit. This will create two items, the original file without resource fork and a directory of the same name with three trailing periods containing the resources. Be sure to move them together to avoid problems. You need to use /usr/Carbon/bin/LaunchCFMApp tool from the Unix command line." This file contains all sorts of useful information, but one thing that it doesn't mention is that if your app links against any non-System shared libs, you'll have to run them through the UFS Converter in addition to putting them in the "PEFLibraries" folder. Later, _______________________________________________________________________ Matt Henderson metrowerks Corporation CodeWarrior IDE Minion Voice: (512) 873-4700 mhenders@metrowerks.com Fax: (512) 873-4900 From sanguish at digifix.com Wed May 19 14:05:24 1999 From: sanguish at digifix.com (Scott Anguish) Date: Thu Nov 3 14:42:04 2005 Subject: AppleScripting from OSXS? In-Reply-To: <199905191631.KAA26985@floyd.stone.com> References: <199905191631.KAA26985@floyd.stone.com> Message-ID: <199905192105.RAA10343@digifix.com> Andrew Stone wrote: > At the Cocoa talks on Friday last week at WWDC, someone mentioned > that they had haqued a way to send AppleEvents from the Yellow Box. > This would be more ideal than testing scriptability from the Blue Box > - since you'd be able to watch as the script executed. > > Whoever that was, would you mind updating us on what you have and > it's availability? I can help with this Paul Suh sent me the link to this.. http://www.gslink.com/~plsuh/YellowBoxScripting/ScriptEditor.sit.hqx I thought it was a YB app, so why it is Stuffed and binhexed I have no idea. From pekeler at codefab.com Wed May 19 14:53:06 1999 From: pekeler at codefab.com (Christian Pekeler) Date: Thu Nov 3 14:42:04 2005 Subject: ANNOUNCE: MacOSXAmp 0.1 is available In-Reply-To: <37431C11.CC7A990C@harmony-ds.com> Message-ID: <9905191753.AA061505@bass> [MacOSXAmp] > The main reason for this release is to get feedback Ok. After looking at http://www.harmony-ds.com/MacOSXAmp/shot1.jpg and http://www.harmony-ds.com/MacOSXAmp/shot2.jpg I can't stop myself from ranting. This application like so many other sound-related applications ignores all GUI standards. I was a great fan of NEXTSTEP because nearly all applications look similar and consistent. This is a great advantage everyone should understand. As far as I know the MacOS it has the same goal and that's what the Mac is famous for. There is no conceptional difference between a CDPlayer for instance and other applications like mailviewer which justifies that the user has to learn yet another GUI. If the programmer wants to show her creativity, then do a fancy info panel. But please don't experiment with the main gui. If I set that the main backgroundcolor of windows should be blue instead of grey in the system's preferences, I want ALL applications to apply to my settings. If I change the system font to a bigger size because I'm visually impaired, I want ALL applications to apply to my settings. If Apple updates the AppKit to change the appearance of buttons, I want all applications to have this new type of button. Will that be possible with MacOSXAmp's custom gui? I don't care if Windows or X-windows applications invent their own guis. But please don't do this on my favorite platform. I think every application developer has his share of responsibility how the public perceives the operating system as a whole. Please take a look at Apple/Documentation/Developer/YellowBox/ReleaseNotes/AppLayout.html for a start. Sorry, but I had to let this out. Christian From malcolm at saracen.plsys.co.uk Tue May 18 15:23:05 1999 From: malcolm at saracen.plsys.co.uk (mmalcolm crawford) Date: Thu Nov 3 14:42:04 2005 Subject: ANNOUNCE: MacOSXAmp 0.1 is available Message-ID: <199905182223.XAA00605@saracen.plsys.co.uk> > Ok. After looking at http://www.harmony-ds.com/MacOSXAmp/shot1.jpg and > http://www.harmony-ds.com/MacOSXAmp/shot2.jpg I can't stop myself from > ranting. This application like so many other sound-related applications > ignores all GUI standards. I was a great fan of NEXTSTEP because nearly all > applications look similar and consistent. > Oh heavens, I'm sorry but I'm moved to write a "me too" follow-up here. Creating consistent UIs is just so easy using AppKit, it's actually more difficult to do something different. I said this to the Apple Cork team when they were here a couple of years ago (gosh, is it really that long?! Apologies to any of them reading this.) for training prior to writing Stickies.app -- I even put in an extra page of notes about UI consistency. I know Stickies may be in a slightlydifferent category since it has some "historical sentiment", but even so... Follow-ups perhaps to -talk? Best wishes, mmalc. From jcr at idiom.com Wed May 19 16:33:01 1999 From: jcr at idiom.com (John C. Randolph) Date: Thu Nov 3 14:42:04 2005 Subject: TimeSpans and Time Selections Message-ID: <37434A1D.3C8666F1@idiom.com> Hi all, I'm working on a pair of classes that some might find useful, so I'm calling for comments here. The NSDate classes do a very good job of dealing with instants of time, but there's nothing in the Foundation framework that really deals with Spans of time. What I've written is a time span class which has two dates. The earlier of the two is the start date, the later is the end date. The duration of a time span is zero or positive. NSDistantFuture and NSDistantPast are both valid, so a time span can represent any contiguous period of time. The time selection class is an ordered, self-normalizing list of time spans, which can represent discontinuous spans of time. Any addition of a time span to a time selection will cause the selection to re-normalize itself. So, the part I'm mulling over, is whether to attempt "every other tuesday" type of functionality, and also dependent timespans (e.g. "starting a week after that other time span's end date.") Right now time spans and time selections are all expressed in terms of absolute dates. This code will be released, most likely to the MiscKit. Comments? -jcr From asen at dti.net Wed May 19 19:08:22 1999 From: asen at dti.net (Anders Sensenig) Date: Thu Nov 3 14:42:04 2005 Subject: Query By Example Message-ID: <199905200209.WAA01321@localhost> Hello, Could someone give me a quick primer on implementing a QBE in code, say for an NSString entity? A pointer to an example the dev examples would be great, too. Thanks. -- Anders -- From wchin at acm.org Wed May 19 22:32:01 1999 From: wchin at acm.org (Bill Chin) Date: Thu Nov 3 14:42:04 2005 Subject: Samba dbench and MOXS profiling Message-ID: <199905200435.AAA20250@mail.richmond.net> Ken Case pointed out that gprof only measures CPU time spent in the process and not time spent in the kernel. He suggested using Sampler, and I did. The results are below. Basically, dbench is spending almost half its time in _open(). Time spent in read() and write() is relatively small. That means to me that disk throughput is pretty good, but the open() call is sabotaging performance. That means one has to dig into the Darwin source to figure out why it's this slow. Does anyone know of an equivalent to Sampler for Linux, specifically on Alpha? I'm going to work on the file locking issues first anyways... they are much more important than performance tuning right now. Sampling of /Network/Users/bchin/Library/OpenUp/dbench_tgz_0/dbench/dbench - #3 Sampling at 1999-05-20 01:17:14 -0400 Samples (displayed/total): 755/755 Call graph: start 753 __start 753 _main 753 _do_one_child 753 _child_run 753 _do_open 326 _open 324 _open [STACK TOP] 324 _strupper 2 _toupper 2 error_message 1 error_message [STACK TOP] 1 _toupper 1 _toupper [STACK TOP] 1 _do_unlink 159 _unlink 158 _unlink [STACK TOP] 158 _strupper 1 _toupper 1 _toupper 1 mcount 1 mcount [STACK TOP] 1 _do_write 108 _write 107 _write [STACK TOP] 107 _lseek 1 _lseek [STACK TOP] 1 _do_rename 78 _rename 78 _rename [STACK TOP] 78 _do_read 23 _read 22 _read [STACK TOP] 22 _lseek 1 _lseek [STACK TOP] 1 _all_string_sub 20 _strstr 12 _strstr 7 _strstr [STACK TOP] 5 mcount 2 mcount 2 _moncount 2 _moncount [STACK TOP] 2 _strncmp 3 _strncmp [STACK TOP] 3 _strlen 2 _strlen [STACK TOP] 2 _all_string_sub 5 mcount 5 mcount 3 _moncount 3 _moncount [STACK TOP] 3 0xbffff730 2 _moncount 2 _moncount [STACK TOP] 2 _strlen 3 _strlen [STACK TOP] 3 _strtok 11 _strtok 10 _strtok [STACK TOP] 6 mcount 4 __mh_execute_header 2 _moncount 2 _moncount [STACK TOP] 2 0xbffff740 1 _moncount 1 _moncount [STACK TOP] 1 mcount [STACK TOP] 1 _strtok [STACK TOP] 1 _do_mkdir 11 _mkdir 11 _mkdir [STACK TOP] 11 _do_close 5 _close 5 _close [STACK TOP] 5 _fgets 5 ___srefill 2 ___sread 2 _read 2 _read [STACK TOP] 2 _memcpy 2 _memcpy 2 mcount 1 __mh_execute_header 1 _moncount 1 _moncount [STACK TOP] 1 _memcpy [STACK TOP] 1 _fgets 1 _fgets [STACK TOP] 1 _do_stat 3 _strupper 2 _toupper 2 _toupper 2 mcount 1 0xbffff610 1 _moncount 1 _moncount [STACK TOP] 1 _toupper [STACK TOP] 1 _stat 1 _stat [STACK TOP] 1 _atoi 3 _strtol 2 mcount 1 mcount [STACK TOP] 1 _strtol 1 _strtol [STACK TOP] 1 0xbffff770 1 _dequeue_unlock_send 1 _dequeue_unlock_send [STACK TOP] 1 _child_run 1 _child_run [STACK TOP] 1 0x411135d0 1 0x41100888 1 0x41104c6c 1 0x41108b1c 1 0x41107938 1 0x4111bc7c 1 0x4111bc5c 1 0x4111bc5c [STACK TOP] 1 0xbfffff00 1 0x4111356c 1 0x4111356c [STACK TOP] 1 Total number in stack (recursive counted multiple, when >=5): 9 mcount 7 _moncount 6 _toupper Sort by top of stack, same collapsed (when >= 5): _open [STACK TOP] 324 _unlink [STACK TOP] 158 _write [STACK TOP] 107 _rename [STACK TOP] 78 _read [STACK TOP] 24 _moncount [STACK TOP] 12 _mkdir [STACK TOP] 11 _strtok [STACK TOP] 7 _strstr [STACK TOP] 5 _close [STACK TOP] 5 _strlen [STACK TOP] 5 -- Bill Chin (NeXTmail/MIME accepted) From raoul at mobopro.com Wed May 19 22:56:13 1999 From: raoul at mobopro.com (Raoul Duke) Date: Thu Nov 3 14:42:04 2005 Subject: (SUMMARY) Not a concrete instance?? References: <3733750E.6D9152F2@mobopro.com> Message-ID: <3743A3FD.931B2FC6@mobopro.com> Raoul Duke wrote: > > I have a question so basic that I'm lost. When I invoke this line: > > Column *col = [[Column alloc] initWithCapacity:rows]; [snip] > @interface Column:NSMutableArray This should likely be in the FAQ (if there is such a thing), given that apparently it's been on the list a few times. The short answer: You can't simply subclass NSMutableArray (and other class clusters, like NSArray, NSString, and NSDictionary) because they are not definite (concrete) classes. To get the desired effect, it's easiest to include an instance variable in the subclass and reimplement those methods you need without counting on [super ...] calls. That is, my Column class would instead be: @interface Column : NSObject { float width; NSMutableArray internalArray; } - initWithCapacity:(unsigned int)numRows; - (void)setWidth:(float)aWidth; - (float)width; // Declaration of all the (needed) array access methods... - (void) addObject:(id)anObject; [...] - initWithCapacity:(unsigned int)numRows { self = [super init]; internalArray = [[NSMutableArray arrayWithCapacity:numRows] retain]; width = DEFAULT_WIDTH; return self; } --Raoul From nat at object-factory.com Wed May 19 23:21:09 1999 From: nat at object-factory.com (Georg Wallmann) Date: Thu Nov 3 14:42:04 2005 Subject: ANNOUNCE: MacOSXAmp 0.1 is available Message-ID: <199905200621.IAA00746@walitza.object-factory.com> Yeah, I hate the non-standard GUI elements of Quicktime 4 too... Nat! P.S. http://www.harmony-ds.com/MacOSXAmp/shot2.jpg looks sweet IMO. Since you can apparently change everything in a skin, it shouldn't be feasible to create a "compliant" skin. From asen at dti.net Thu May 20 06:59:40 1999 From: asen at dti.net (Anders Sensenig) Date: Thu Nov 3 14:42:04 2005 Subject: Query By Example Message-ID: <199905201400.KAA01602@localhost> Hello, To be more specific: I know how to create an EOAssociation between a display object and the .query= attribute. However, what I want to do is to pass the query values and fetch entirely programmatically. I'm looking at queryBindingValues and queryOperatorValues (and their corresponding set methods) as possible applicable methods, but I'm not sure. For instance, queryBindingValues has the following documentation: Returns a dictionary containing the actual values that the user wants to query upon. You use this method to perform a query stored in the model file. Bind keys in this dictionary to elements on your component that specify query values, then pass this dictionary to the fetch specification that performs the fetch. Any hints on implementing this? -- Anders > Could someone give me a quick primer on implementing a QBE > in code, say for an NSString entity? A pointer to an > example the dev examples would be great, too. Thanks. > From maury at OAAI.COM Thu May 20 07:20:30 1999 From: maury at OAAI.COM (Maury Markowitz) Date: Thu Nov 3 14:42:04 2005 Subject: Sending mouse events programatically Message-ID: <199905201416.KAA20205@OAAI.COM> > When creating mouse click events, how does one generate a valid event > number? If I take the current event number and add 1 to it, will that upset > the next event to be created by the system? Could this be the problem? This definitely seems to be a problem, although perhaps only one of a number. I was recently trying to do the same thing to allow for "multi-drags" in our app, but was unable to sythesize the proper events - which is that the d&d code needs. Every attempt of mine resulted in errors related to sequencing, I even tried trapping the last mouseDown and feeding that back in, with no luck. Maury From Steve.Ivy at cbn.org Thu May 20 07:49:29 1999 From: Steve.Ivy at cbn.org (Ivy, Steven) Date: Thu Nov 3 14:42:04 2005 Subject: Inheriting a subclass Message-ID: <2FAD21AE9E37D21191F900805FE6778732203E@NT2> I've been playing around with NSOutlineView, and while doing research on implementations, ran across Andrew Stone's excellent SDTableView code (http://www.omnigroup.com/MailArchive/MacOSX-dev/1998/2926.html). What I'm wondering is if I want an NSOutlineView (which inherits from NSTableView) to inherit behavior from SDTableView, can that be done? And how? Thanks- --Steve ---------------------------------------------------- "Should we talk about the weather? Should we talk about the government?" -REM Steve Ivy, steve.ivy@cbn.org ---------------------------------------------------- From leigh at ernst.tomandandy.com Thu May 20 15:18:13 1999 From: leigh at ernst.tomandandy.com (Leigh Smith) Date: Thu Nov 3 14:42:04 2005 Subject: teTeX - anyone? Message-ID: <199905202218.SAA06246@ernst.tomandandy.com> I ported the most recent teTeX 0.9 to MOXS and its pretty straightforward. Thomas Esser now has the changes necessary to have it compile out of the box (gcc had most of the __unix__ definitions removed), as well as a bug fix for filenames in dvips. teTeX 0.9 includes pdfTeX, but for now I use the postscript CM fonts and psprepare to produce a viewable .ps file with Preview. This is actually a pretty neat solution, if you use make to run dvips through the dvi file automatically. Beware the latest dvips by default does not turn on the -K option to throw away included .ps comments, which caused me some head scratching. To produce a pdf file you will need postscript CM fonts anyway. While TeXView would be nice, it is not necessary to do production level TeX stuff. -- Leigh Smith leigh@tomandandy.com (MIME) tomandandy +1-212-334-0421 (W) +1-212-334-0422 (F) 89 Greene St. New York, NY 10012, USA http://www.cs.uwa.edu.au/~leigh Microsoft - What do you want to re-install today? From portnoy at ai.mit.edu Thu May 20 16:09:43 1999 From: portnoy at ai.mit.edu (Stephen Peters) Date: Thu Nov 3 14:42:04 2005 Subject: teTeX - anyone? In-Reply-To: Leigh Smith's message of "Thu, 20 May 1999 15:34:33 -0700 (PDT)" References: <199905202218.SAA06246@ernst.tomandandy.com> Message-ID: Leigh Smith writes: > I ported the most recent teTeX 0.9 to MOXS and its pretty > straightforward. Yep. If people are interested, I'm going to try to put a binary Installer package and put it on peak. People interested in trying my non-Installer binary now can go to http://www.portnoy.org/rhapsody/ and read the instructions there. > teTeX 0.9 includes pdfTeX, but for now I use the postscript CM fonts > and psprepare to produce a viewable .ps file with Preview. This is > actually a pretty neat solution, if you use make to run dvips > through the dvi file automatically. It would be nicer if Preview had a way to re-load the document. :-) > To produce a pdf file you will need postscript CM fonts > anyway. While TeXView would be nice, it is not necessary to do > production level TeX stuff. Actually, the neat bit is that the texmf library actually *has* those fonts (bluesky's Type 1), and pdftex includes them when it creates a PDF file. It's gorgeous -- I may never go back to DVI :-) -- Stephen L. Peters portnoy@ai.mit.edu PGP fingerprint: BFA4 D0CF 8925 08AE 0CA5 CCDD 343D 6AC6 "Poodle: The other white meat." -- Sherman, Sherman's Lagoon From rcfa at cubiculum.com Thu May 20 15:51:30 1999 From: rcfa at cubiculum.com (Ronald C.F. Antony) Date: Thu Nov 3 14:42:04 2005 Subject: Inheriting a subclass In-Reply-To: <2FAD21AE9E37D21191F900805FE6778732203E@NT2> References: <2FAD21AE9E37D21191F900805FE6778732203E@NT2> Message-ID: <199905202251.SAA27215@kannix.cubiculum.com> you wrote: > I've been playing around with NSOutlineView, and while doing research on > implementations, ran across Andrew Stone's excellent SDTableView code > ( > What I'm wondering is if I want an NSOutlineView (which inherits from > NSTableView) to inherit behavior from SDTableView, can that be done? And > how? Not without source, since you don't know what assumptions NSOutlineView makes about its superclass. Possibly you could let SDTableView poseAs: NSTableView, but if that's possible depends on if they have the same size. Lastly you could do runtime hacks, like modifying pointers to the superclass, but a hack is a hack, and the results are potentially disasterous. Ronald ============================================================================== "The reasonable man adapts himself to the world; the unreasonable one persists in trying to adapt the world to himself. Therefore all progress depends on the unreasonable man." G.B. Shaw | rcfa@cubiculum.com | NeXT-mail welcome From brianw at irisgroup.com Thu May 20 19:44:33 1999 From: brianw at irisgroup.com (Brian Wagner) Date: Thu Nov 3 14:42:04 2005 Subject: JFIF & GIF, can they be created? Message-ID: <3744C890.5E8235BE@irisgroup.com> I'm trying to create a program that can take text and draw it to a GIF or JFIF file. I may have bitten off more than I can chew for learning OSX development but I've done it on a Mac already. I'm trying to do it with Java as Objective -C looks too tough. I'm seeing several obsticles. Since OSX's Java can't create image files I have to do it with the Java-Yellow box classes. (Is Java 2 coming soon?) But, the Java browser doesn't have NSString so how do I get text into an NSView? Can I even create a GIF or JFIF? From the newsgroup it looks like JFIF is out. Any insight, opinions or plain old help it appreciated in advance. From inphlux at lotek.org Thu May 20 22:44:45 1999 From: inphlux at lotek.org (ix) Date: Thu Nov 3 14:42:04 2005 Subject: MacOSXAmp 0.1 Message-ID: <199905210504.BAA06506@project.lotek.org> > P.S. http://www.harmony-ds.com/MacOSXAmp/shot2.jpg looks > sweet IMO. Since you can apparently change everything in a skin, it > shouldn't be feasible to create a "compliant" skin. just curious. has anyone actually gotten this app to work? when i click on the .app in workspace manager, nothing happens. i also have had similarly bad luck with the other mp3 players. mpap.app launches but doesnt play, and omniamp gave similar results. im guessing it may have something to do with nssound replacing soundkit. system in question is g3/300 without av hardware running beaker1N5. thanks in advance for any tips/advice. From raoul at mobopro.com Thu May 20 23:11:44 1999 From: raoul at mobopro.com (Raoul Duke) Date: Thu Nov 3 14:42:04 2005 Subject: App name Message-ID: <3744F920.5B61C1EC@mobopro.com> Quick question--I didn't see this in the docs. Is there a neat way of getting an application's name programmatically--say, Foo for a class executing as part of Foo.app? I guess [[[[NSBundle mainBundle] bundlePath] lastPathComponent] stringByDeletingPathExtension] would do it, but that's a bit gross. --Raoul From sebestyen_g at accentcomm.com Fri May 21 00:03:09 1999 From: sebestyen_g at accentcomm.com (Gabor Sebestyen) Date: Thu Nov 3 14:42:04 2005 Subject: Uname for OPENSTEP? Message-ID: <3745052D.474268FD@accentcomm.com> Does uname exist for OPENSTEP? Best regards, Gabor Sebestyen -------------------------------------------------------------- A C C E N T C O M M U N I C A T I O N S - We Make Your Web Groove - Tel: (+36 1) 315-0701 http://www.accentcomm.com (+36 1) 343-4440 e-mail: info@accentcomm.com GSM: (+36 30) 951-4334 Fax: (+36 30) 980-0344 -------------------------------------------------------------- From cmh at greendragon.com Fri May 21 00:43:02 1999 From: cmh at greendragon.com (Chris Hanson) Date: Thu Nov 3 14:42:04 2005 Subject: DriverKit questions Message-ID: Now that DriverKit is available on the Darwin site, I was wondering if anyone would share their experiences using it to develop drivers for Mac OS X Server. I realize that DriverKit is going away soon in favor of IOKit, but that doesn't make me less curious... :) Here's what I'm particularly curious about: 1. Is everything necessary to build drivers that will work with the currently shipping version of Mac OS X Server included with the Darwin DriverKit? 2. How easy is it to bring up a random PCI card for which you have specs, and for which a suitable base class already exists? (For instance, a SCSI, video, or Ethernet card.) 3. Do PCI cards have to be recognized by Open Firmware in order for DriverKit to load their drivers? (I don't even know if it's possible to have a PCI card not be recognized by OF...) 4. How hard is it to bring up a random PCI card for which you have specs but for which a suitable base class does not exist? (For instance, an analog or digital I/O card of some sort, or a TV tuner card, or a PCMCIA adaptor card?) -- Chris From norbert.heger at obdev.at Fri May 21 00:50:20 1999 From: norbert.heger at obdev.at (Norbert Heger) Date: Thu Nov 3 14:42:04 2005 Subject: App name References: <3744F920.5B61C1EC@mobopro.com> Message-ID: <9905210750.AA00502@cray> Raoul Duke wrote: > Quick question--I didn't see this in the docs. Is there a neat > way of getting an application's name programmatically--say, Foo > for a class executing as part of Foo.app? Quick answer -- Try [[NSProcessInfo processInfo] processName] Norbert _____________________________ _________________________________ O B J E C T I V E D E V E L O P M E N T Norbert Heger OpenStep MacOS-X EOF WebObjects norbert.heger@obdev.at Software Development +43-664-2811777 www.obdev.at From Axel.Katerbau at gmx.net Fri May 21 01:36:33 1999 From: Axel.Katerbau at gmx.net (Axel Katerbau) Date: Thu Nov 3 14:42:04 2005 Subject: teTeX - anyone? (OmniPDF) Message-ID: <199905210837.KAA18731@nereus.informatik.uni-bonn.de> > It would be nicer if Preview had a way to re-load the document. Also OmniPDF would be nicer with a reload functionality. An "open foo.pdf" command could reload if the file is newer. That would be really great. Axel From jes at nexus.rednsi.com Fri May 21 02:31:22 1999 From: jes at nexus.rednsi.com (Josep Egea i Sanchez) Date: Thu Nov 3 14:42:04 2005 Subject: NSBox subclass in an IB palette Message-ID: <199905211021.MAA31766@filemon.rednsi.com> Hi, I've created an NSBox subclass that merely changes the way it looks. I've created a new IB palette and added an instance of this class to it. When I try to use the object in this palette, it mostly works (it draws correctly, I can move, resize and inspect it in IB without problems) but IB doesn't treat it as an NSBox. I cannot insert other views inside it and if I double-click over it, it seems to work well (its handles are partly overlapped as it happens with a real NSBox) but if i try to drop any other object inside it, the object ends up in the parent view. Does anybody know how to make this work? I've tried to create a "dumb" subclass of NSBox (one with no new methods or instance variables) just to see if one of my custom methods was interfering with IB, but this class suffers from the same problem. To me, it looks like IB has NSBox's class name "hard-coded" somewhere and refuses to work with my subclasses. Any help would be greatly appreciated. PS: I'm using OPENSTEP for Mach 4.2 Intel + EOF 2.2 Best regards --- Josep Egea - jes@rednsi.com - NeXTMail & MIME OK NEXUS Servicios de Informacion - Barcelona (Spain) From lavoie at cst.ca Fri May 21 04:39:17 1999 From: lavoie at cst.ca (Martin-Gilles) Date: Thu Nov 3 14:42:04 2005 Subject: MacOSXAmp 0.1 Message-ID: <199905211139.HAA13161@plexus.cst.ca> >> P.S. http://www.harmony-ds.com/MacOSXAmp/shot2.jpg looks >> sweet IMO. Since you can apparently change everything in a skin, it >> shouldn't be feasible to create a "compliant" skin. > >just curious. has anyone actually gotten this app to work? when i click on >the .app in workspace manager, nothing happens. i also have had similarly >bad luck with the other mp3 players. mpap.app launches but doesnt play, and >omniamp gave similar results. im guessing it may have something to do with >nssound replacing soundkit. system in question is g3/300 without av hardware >running beaker1N5. thanks in advance for any tips/advice. Works fine on my stock 8600/200, and on a PowerForce G3-upgraded 8100 @ 300Mghz. =========================================================================== Martin-Gilles Lavoie, Mac OS programmer, Corporate Software & Technologies --------------------------------------------------------------------------- (internet) http://pages.infinit.net/mouser/ mouser@videotron.ca (intranet) http://blackhole.cst.ca/ lavoie@cst.ca --------------------------------------------------------------------------- "No! Try not. Do, or do not. There is no try." -- Yoda on error handling From lavoie at cst.ca Fri May 21 05:58:14 1999 From: lavoie at cst.ca (Martin-Gilles) Date: Thu Nov 3 14:42:04 2005 Subject: MacOSXAmp 0.1 Message-ID: <199905211258.IAA14694@plexus.cst.ca> >Works fine on my stock 8600/200, and on a PowerForce G3-upgraded 8100 @ >300Mghz. That's G3-upgraded 8500/200. ^ My mistake. =========================================================================== Martin-Gilles Lavoie, Mac OS programmer, Corporate Software & Technologies --------------------------------------------------------------------------- (internet) http://pages.infinit.net/mouser/ mouser@videotron.ca (intranet) http://blackhole.cst.ca/ lavoie@cst.ca --------------------------------------------------------------------------- "No! Try not. Do, or do not. There is no try." -- Yoda on error handling From epeyton at epicware.com Fri May 21 06:22:14 1999 From: epeyton at epicware.com (Eric Peyton) Date: Thu Nov 3 14:42:04 2005 Subject: MacOSXAmp 0.1 In-Reply-To: <199905210504.BAA06506@project.lotek.org> Message-ID: Check out the tip on www.epicware.com/Tips about how to set up soundkit to work correctly under Mac OS X Server ... Eric P.S. Yes this is a gratuitous plug :) On Thu, 20 May 1999, ix wrote: > > P.S. http://www.harmony-ds.com/MacOSXAmp/shot2.jpg looks > > sweet IMO. Since you can apparently change everything in a skin, it > > shouldn't be feasible to create a "compliant" skin. > > just curious. has anyone actually gotten this app to work? when i click on > the .app in workspace manager, nothing happens. i also have had similarly > bad luck with the other mp3 players. mpap.app launches but doesnt play, and > omniamp gave similar results. im guessing it may have something to do with > nssound replacing soundkit. system in question is g3/300 without av hardware > running beaker1N5. thanks in advance for any tips/advice. > > -- Eric Peyton epeyton@epicware.com Software and Source for Mac OS X Server From epeyton at epicware.com Fri May 21 06:34:15 1999 From: epeyton at epicware.com (Eric Peyton) Date: Thu Nov 3 14:42:04 2005 Subject: MacOSXAmp Message-ID: 1) Dev group should ignore my last message if not approriate ... this is a talk thing not a dev thing - I am sorry. (Never hit reply all :-( ) 2) ix - those tips might not help you at all - my fault. Is there any console output (Click on the Workspace, Hit Cmd-Sh-C) after you run the app. Possible saying sound mode not suppported or some other? Do you get sound in other applications (MailViewer, ProjectBuilder, etc.). I'm not familiar with the older hardware, but that may be your problem. It may not have a sound mode that is supported. Have you gone to the sound preferences panel and made sure your sound wasn't muted? Just some tips ... Eric On Thu, 20 May 1999, ix wrote: > > P.S. http://www.harmony-ds.com/MacOSXAmp/shot2.jpg looks > > sweet IMO. Since you can apparently change everything in a skin, it > > shouldn't be feasible to create a "compliant" skin. > > just curious. has anyone actually gotten this app to work? when i click on > the .app in workspace manager, nothing happens. i also have had similarly > bad luck with the other mp3 players. mpap.app launches but doesnt play, and > omniamp gave similar results. im guessing it may have something to do with > nssound replacing soundkit. system in question is g3/300 without av hardware > running beaker1N5. thanks in advance for any tips/advice. > > -- Eric Peyton epeyton@epicware.com Software and Source for Mac OS X Server From mike at lorax.com Fri May 21 08:40:43 1999 From: mike at lorax.com (Mike Ferris) Date: Thu Nov 3 14:42:04 2005 Subject: RTF data and NSTextView Message-ID: <199905211636.JAA01269@boom.lorax.com> Probably the discrepancy here is in the line fragment padding used by field editors versus the line fragment padding used by default NSTextView systems. I would recommend using the same strategy to display and edit to avoid small differences in the setup of the text systems. If you use the field editor to edit, you might consider using it to draw as well. If you're going to use a cell to edit, use a cell to display as well. (Using a cell might be your best approach here.) Mike Ferris From mtarbell at tarbell.jpl.nasa.gov Fri May 21 08:45:35 1999 From: mtarbell at tarbell.jpl.nasa.gov (Mark Tarbell) Date: Thu Nov 3 14:42:04 2005 Subject: teTeX - anyone? (OmniPDF) In-Reply-To: <199905210837.KAA18731@nereus.informatik.uni-bonn.de> References: <199905210837.KAA18731@nereus.informatik.uni-bonn.de> Message-ID: <199905211545.IAA01032@tarbell.jpl.nasa.gov> > Also OmniPDF would be nicer with a reload functionality. An "open > foo.pdf" command could reload if the file is newer. Is there any way to use OmniPDF (or other app) to convert RTF or HTML into PDF? What are people using to generate PDF documents? Mark -- Mark Tarbell (Mark.A.Tarbell@jpl.nasa.gov) Applications Development Section Information Systems Development and Operations Division Engineering and Science Directorate Jet Propulsion Laboratory 4800 Oak Grove Drive MS 502-500 Pasadena, CA 91109 Opinions expressed herein do not reflect those of JPL. From sbrandon at music.gla.ac.uk Fri May 21 08:50:35 1999 From: sbrandon at music.gla.ac.uk (Stephen Brandon - SysAdmin) Date: Thu Nov 3 14:42:04 2005 Subject: RTF data and NSTextView In-Reply-To: <199905211636.JAA01269@boom.lorax.com> References: <199905211636.JAA01269@boom.lorax.com> Message-ID: <199905211550.QAA14495@music.gla.ac.uk> Mike Ferris wrote: > Probably the discrepancy here is in the line fragment padding used > by field editors versus the line fragment padding used by default > NSTextView systems. > > I would recommend using the same strategy to display and edit to > avoid small differences in the setup of the text systems. If you use > the field editor to edit, you might consider using it to draw as > well. If you're going to use a cell to edit, use a cell to display > as well. (Using a cell might be your best approach here.) > > Mike Ferris Hi, Good thoughts there, thanks Mike. The archived rtf was from NeXTSTEP days (Text object) though, so I do want to convert it, and not continue to display using an old Text object (which is gone in MACOSX anyway). What I have ended up doing is some trial and error with my NS app and my OS app side by side, tweaking offset values for the NSTextView object until I got it right. When taking archived rtf (from Text object) and putting into NSTextView object, I subtract 3 from the origin of the frame rect, and add about 5 to the width. This seems to make everything wrap identically, and placement comes out identical on the page. The field editor is just an NSTextView, so is exactly the same as the NSTextView that I'm using for display. Though I had had some problems with the field editor and the NSTextView displaying differently, I found that that was to do with another bug in my app. I fixed that and now everything works well. Stephen From maury at OAAI.COM Fri May 21 09:38:54 1999 From: maury at OAAI.COM (Maury Markowitz) Date: Thu Nov 3 14:42:04 2005 Subject: RTF data and NSTextView Message-ID: <199905211634.MAA22120@OAAI.COM> > using an old Text object (which is gone in MACOSX anyway). What I have ended > up doing is some trial and error with my NS app and my OS app side by side, > tweaking offset values for the NSTextView object until I got it right. This works well until you want to support zooming, so be aware of this issue. And not to disagree with Mike, but an alternate possible cause for the problem is the difference between the font metrics in the two views? I'm pretty sure that's why we get the one-pixel wiggle in GlyphiX when you start up the editor, it appears not to be the view itself, but the text baseline. Maury From portnoy at ai.mit.edu Fri May 21 11:07:38 1999 From: portnoy at ai.mit.edu (Stephen Peters) Date: Thu Nov 3 14:42:04 2005 Subject: teTeX - anyone? (OmniPDF) In-Reply-To: Mark Tarbell's message of "Fri, 21 May 1999 09:08:49 -0700 (PDT)" References: <199905211545.IAA01032@tarbell.jpl.nasa.gov> Message-ID: Mark Tarbell writes: > > Also OmniPDF would be nicer with a reload functionality. An "open > > foo.pdf" command could reload if the file is newer. > > Is there any way to use OmniPDF (or other app) to convert RTF or > HTML into PDF? What are people using to generate PDF documents? I know of no apps which convert RTF or HTML into PDF on MacOS X Server. I imagine on MacOS X (client) it will be native functionality (or at least a part of any RTF or HTML reader), but currently there's no way to do it. What I tended to use to generate PDF documents was Linux or Windows :-) I would generate PostScript files on Rhapsody, and then either use the Linux pstopdf command or Acrobat Distiller to create the PDF output. -- Stephen L. Peters portnoy@ai.mit.edu PGP fingerprint: BFA4 D0CF 8925 08AE 0CA5 CCDD 343D 6AC6 "Poodle: The other white meat." -- Sherman, Sherman's Lagoon From andrew at stone.com Fri May 21 11:32:59 1999 From: andrew at stone.com (Andrew Stone) Date: Thu Nov 3 14:42:04 2005 Subject: RTF data and NSTextView Message-ID: <199905211835.MAA02965@floyd.stone.com> Don't zoom your graphicView - zoom the clipView - and you don't have to worry about any of this... > From: Maury Markowitz > Date: 1999-05-21 11:10:02 -0600 > To: Multiple recipients of list > Subject: Re: RTF data and NSTextView > > This works well until you want to support zooming, so be aware of > this issue. > > And not to disagree with Mike, but an alternate possible cause for > the problem is the difference between the font metrics in the two > views? I'm pretty sure that's why we get the one-pixel wiggle in > GlyphiX when you start up the editor, it appears not to be the view > itself, but the text baseline. > > Maury > > +++ Andrew Stone Fax: 505-345-3424 From Stefan.Boehringer at ruhr-uni-bochum.de Fri May 21 11:42:29 1999 From: Stefan.Boehringer at ruhr-uni-bochum.de (Stefan.Boehringer@ruhr-uni-bochum.de) Date: Thu Nov 3 14:42:04 2005 Subject: Darwin, 4.1 & VM-Code Message-ID: <9905211900.AA00438@christine> A non-text attachment was scrubbed... Name: not available Type: text/enriched Size: 273 bytes Desc: not available Url : /mailman/archive/macosx-dev/attachments/19990521/ed576a1d/attachment.bin From Steve.Ivy at cbn.org Fri May 21 12:06:33 1999 From: Steve.Ivy at cbn.org (Ivy, Steven) Date: Thu Nov 3 14:42:04 2005 Subject: scgi's XFS + Darwin's VFS? Message-ID: <2FAD21AE9E37D21191F900805FE67787322047@NT2> sgi is releasing the source code for XFS - thier high-volume-high-performance-high-bitwidth filesystem. Could this filesystem be conceivably(sp?) layered within Darwin/OSX's Virtual File System? --Steve ---------------------------------------------------- "Should we talk about the weather? Should we talk about the government?" -REM Steve Ivy, steve.ivy@cbn.org ---------------------------------------------------- From fan at chem.ucla.edu Fri May 21 12:15:19 1999 From: fan at chem.ucla.edu (Paul S. Fan) Date: Thu Nov 3 14:42:04 2005 Subject: teTeX - anyone? (OmniPDF) Message-ID: <199905211915.MAA19447@krypton.chem.ucla.edu> > I would generate PostScript files on Rhapsody, and then either > use the Linux pstopdf command or Acrobat Distiller to create the PDF > output. Of course you can also use the Acrobat Distiller in your blue box to convert postscript files to pdf files and then drag them back to your OSX partition. Best, Paul From annard at webwarecorp.com Fri May 21 12:25:17 1999 From: annard at webwarecorp.com (Annard Brouwer) Date: Thu Nov 3 14:42:04 2005 Subject: Darwin, 4.1 & VM-Code In-Reply-To: <9905211900.AA00438@christine> Message-ID: <9905211225.AA171102@facade> Stefan.Boehringer@ruhr-uni-bochum.de wrote: > now that Darwin is out some subtleties in VM-Code may be elucidated. However I'm > interested in Openstep4.1 primarily. Does anybody know which (if any) changes went > into the VM-Code comparing the two mach-kernels? > I asked Chris Kane about this at the conference and one of the things that they hadn't fixed yet was the slinking of a large VM file. Which was too bad. They are aware of it and should be fixed in a next rrelease. Annard From kc at omnigroup.com Fri May 21 13:33:48 1999 From: kc at omnigroup.com (Ken Case) Date: Thu Nov 3 14:42:04 2005 Subject: DriverKit questions Message-ID: <199905212033.NAA24578@ignem.omnigroup.com> > 1. Is everything necessary to build drivers that will work with > the currently shipping version of Mac OS X Server included with > the Darwin DriverKit? Everything necessary is included with Mac OS X Server (if you install the developer packages), it's not necessary to download Darwin. (If you were wondering if it's possible to do with just Darwin and not X Server: possibly, but I haven't actually tried.) > 2. How easy is it to bring up a random PCI card for which you > have specs, and for which a suitable base class already exists? > (For instance, a SCSI, video, or Ethernet card.) Huh! I don't think I've ever done this, actually: we've only written drivers for things which didn't have a suitable base class. There used to be a lot of DriverKit examples, and I know they included sample Video and Ethernet drivers. I'm not sure how easy they'd be to find now, though: maybe if you dig up a copy of NEXTSTEP 3.x Developer... > 3. Do PCI cards have to be recognized by Open Firmware in order > for DriverKit to load their drivers? (I don't even know if it's > possible to have a PCI card not be recognized by OF...) The PC Voodoo2 card I'm using right now on my B&W G3 certainly doesn't have an Open Firmware driver, so I'd say no. :) (Presumably OpenFirmware does detect that some card is present, it just ignores it.) The card _would_ presumably need Open Firmware drivers if it were used during the boot process at all (e.g., if it were a video or SCSI driver). > 4. How hard is it to bring up a random PCI card for which you > have specs but for which a suitable base class does not exist? > (For instance, an analog or digital I/O card of some sort, or a > TV tuner card, or a PCMCIA adaptor card?) Not hard at all, in terms of interfacing with DriverKit: just subclass IODirectDevice (which handles most of the standard PCI configuration) and get to work on your card's specific needs. (Obviously the overall driver complexity depends on the demands of the device in question.) Ken From kc at omnigroup.com Fri May 21 13:33:50 1999 From: kc at omnigroup.com (Ken Case) Date: Thu Nov 3 14:42:04 2005 Subject: Voodoo2-accelerated MesaGL and Glide ports available Message-ID: <199905212033.NAA24589@ignem.omnigroup.com> For those who don't read macosx-admin or macosx-talk, we just released our ports of MesaGL and Glide(tm) (and the Quake 3 Test client, which uses them). For details, see http://www.omnigroup.com/Software/Glide/. Ken From allen at joslin.net Fri May 21 13:42:42 1999 From: allen at joslin.net (allen) Date: Thu Nov 3 14:42:04 2005 Subject: Apache-1 && JServ1.0b5 == NeedHelp! References: <199905211831.LAA25539@ignem.omnigroup.com> Message-ID: <3745C542.F88DDCA2@joslin.net> Apache-1 builds fine (stop at the first compile & sed out the -arch i386's) standard make standard make install Installed JSDK 2.0 into /usr/local/jsdk Then I configure JServ 1.0bx w/this configure command: ./configure \ --host=ppc-apple \ --prefix=/usr/local/jserv \ --with-jsdk=/usr/local/jsdk \ --with-apache-install=/tmp/apache/Release/usr --with-apache-conf That gets it to configure (and admit that apache was built w/DSO) then during 1.0bx's make I get this error: ar cru .libs/libjserv.a ar: no archive members specified usage: ar -d [-TLv] archive file ... ar -m [-TLv] archive file ... ar -m [-abiTLv] position archive file ... ar -p [-TLv] archive [file ...] ar -q [-cTLv] archive file ... ar -r [-cuTLv] archive file ... ar -r [-abciuTLv] position archive file ... ar -t [-TLv] archive [file ...] ar -x [-ouTLv] archive [file ...] make[2]: *** [libjserv.la] Error 1 make[1]: *** [all-recursive] Error 1 make: *** [all-recursive] Error 1 I have been thru the make files & config files and can't understand what to fix or what flags to specify to configure or make to get this one to pass. I really feel like I'm getting closer to a working JServ... Just not close enuf!!! Al; From kc at omnigroup.com Fri May 21 13:57:52 1999 From: kc at omnigroup.com (Ken Case) Date: Thu Nov 3 14:42:04 2005 Subject: scgi's XFS + Darwin's VFS? Message-ID: <199905212057.NAA01417@ignem.omnigroup.com> > sgi is releasing the source code for XFS - thier > high-volume-high-performance-high-bitwidth filesystem. Could this > filesystem be conceivably(sp?) layered within Darwin/OSX's Virtual > File System? Yes, it could. Sounds like a worthwhile project to me. Ken From hunter at lastonepicked.com Fri May 21 14:03:49 1999 From: hunter at lastonepicked.com (Hunter Hillegas) Date: Thu Nov 3 14:42:04 2005 Subject: Voodoo2-accelerated MesaGL and Glide ports available Message-ID: <19990521210227487.AAA483@nts1.webhostingprovider.com@firewall.jacobstern.com> Does Quake II use this as well? Hunter ---------- >From: Ken Case >To: Multiple recipients of list >Subject: Voodoo2-accelerated MesaGL and Glide ports available >Date: Fri, May 21, 1999, 1:53 PM > > For those who don't read macosx-admin or macosx-talk, we just > released our ports of MesaGL and Glide(tm) (and the Quake 3 Test > client, which uses them). > > For details, see http://www.omnigroup.com/Software/Glide/. > > Ken > From bwebster at owlnet.rice.edu Fri May 21 14:30:20 1999 From: bwebster at owlnet.rice.edu (Brian Webster) Date: Thu Nov 3 14:42:04 2005 Subject: Transparency in NSImages Message-ID: <199905212130.QAA07962@marsh.owlnet.rice.edu> I have a program in which I want to draw several images on top of one another, with some of the images having transparent sections so that each new image doesn't completely overwrite what was drawn before, just the non-transparent sections. I first create a new, blank image, and then draw each of the other images on top of it using lockFocus: and compositeToPoint:. But, after trying many different things, I haven't been able to get it to work (the transparent part draws white over the previous image(s)). I created the images using TIFFany, and it assures me that there is indeed a transparent portion of the image. hasAlpha: even returns YES, but when I do the drawing, no cigar. Has anyone dealt with this before? Is there anything I need to fool with with the NSImage or the NSBitmapImageRep that it uses? -- Brian Webster bwebster@owlnet.rice.edu http://www.owlnet.rice.edu/~bwebster/ From bwebster at owlnet.rice.edu Fri May 21 14:22:22 1999 From: bwebster at owlnet.rice.edu (Brian Webster) Date: Thu Nov 3 14:42:04 2005 Subject: Converting to PDF Message-ID: <199905212130.QAA07958@marsh.owlnet.rice.edu> Mark Tarbell writes: > Is there any way to use OmniPDF (or other app) to convert RTF or > HTML into PDF? What are people using to generate PDF documents? There is a version of Ghostscript for the Mac that will convert Postscript to PDF at ftp://ftp.cs.wisc.edu/pub/ghost/aladdin/gs550/mac/ You'd have to use Blue Box and you'd have to save the HTML or RTF to Postscript, but it'd work eventually. ;-) -- Brian Webster bwebster@owlnet.rice.edu http://www.owlnet.rice.edu/~bwebster/ From tpugh at oce.orst.edu Fri May 21 15:12:42 1999 From: tpugh at oce.orst.edu (tpugh@oce.orst.edu) Date: Thu Nov 3 14:42:04 2005 Subject: sgi's XFS + Darwin's VFS? In-Reply-To: <199905212057.NAA01417@ignem.omnigroup.com> Message-ID: <990521151242.399AADoJ.tpugh@rarebear> One last comment before moving this to macosx-talk. I would like to second a port of sgi's XFS filesystem if its true. The XFS filesystem from SGI is excellent! And Apple needs a better file system to replace UFS. Apple needs a "journalled" filesystem to cope with hard drives today and tomorrow. Otherwise, we'll be waiting for hours for our Macs to boot after a system crash, because the system is doing a file system check on my 50GB hard drive. And there is no way around file system check without corrupting your file system. Here is a white paper on the XFS filesystem: http://www.sgi.com/Technology/xfs-whitepaper.html And a teaser: 1.1 Product Feature Summary =========================== XFS provides a full 64-bit filesystem capable of scaling easily to handle extremely large files and filesystems that can grow to 1 terabyte (up to 9 million terabytes in successor releases). XFS's major features include: - full 64-bit file capabilities -large files -large filesystems -large numbers of files -sparse files (files with "holes") - rapid and reliable recovery of filesystem structure using journalling technology - integrated, full-function volume manager called XLV - extremely high I/O performance that scales well on multiprocessing systems - guaranteed rate I/O for multimedia and data acquisition uses - back up of filesystems while still in use, significantly reducing administrative overhead - XFS is compatible with existing applications, EFS filesystems, and NFS Additionally, systems can use XFS filesystems exclusively or have a mixture of XFS and EFS filesystems. 1.2 Background Information =========================== The XFS filesystem is a "journalled" filesystem. This means that updates to filesystem metadata (inodes, directories, bitmaps, etc.) are written to a serial log area on disk before the original disk blocks are updated in place. In the event of a crash, such operations can be redone using data present in the log, to restore the filesystem to a consistent state. XFS includes a fully integrated volume manager, XLV, which creates an abstraction of a sequence of logical disk blocks, called a volume, to be used by the XFS filesystem. The sequence of disk blocks (a volume) can be assembled by concatenating, plexing (mirroring), and striping across a number of physical disk drives or RAID devices. XLV ensures that plexed data is kept consistent across all plexes and will discussed in detail later. - Tim - -------------------------------------------------------------- Tim F. Pugh email: tpugh@oce.orst.edu Oceanic and Atmospheric Sciences voice: 541-737-4414 Oregon State University fax: 541-737-2064 104 Ocean Admin Bldg Corvallis, Oregon 97331-5503 NeXTmail, MIME, Sun, or Ascii mail ok! -------------------------------------------------------------- From kc at omnigroup.com Fri May 21 15:13:25 1999 From: kc at omnigroup.com (Ken Case) Date: Thu Nov 3 14:42:04 2005 Subject: Voodoo2-accelerated MesaGL and Glide ports available Message-ID: <199905212213.PAA24038@ignem.omnigroup.com> [Followups to macosx-talk, please.] > Does Quake II use this [hardware acceleration] as well? It does now. See http://www.omnigroup.com/Software/Quake2/ for details. Ken From maury at OAAI.COM Fri May 21 15:23:49 1999 From: maury at OAAI.COM (Maury Markowitz) Date: Thu Nov 3 14:42:04 2005 Subject: Transparency in NSImages Message-ID: <199905212219.SAA22759@OAAI.COM> > I have a program in which I want to draw several images on top of one > another, with some of the images having transparent sections so that each > new image doesn't completely overwrite what was drawn before Right. > non-transparent sections. I first create a new, blank image, and then draw > each of the other images on top of it using lockFocus: and > compositeToPoint: The problem is that you have to fill the image with "clear paint" first. Here, try this just after you create the image... [image lockFocus]; PSsetgray(NSWhite); PSsetalpha(0.0); NSRectFill([image size]); PSsetalpha(1.0); add your drawing code here. Enjoy! Maury From Malcolm_Crawford at plsys.co.uk Fri May 21 15:25:11 1999 From: Malcolm_Crawford at plsys.co.uk (mmalcolm crawford) Date: Thu Nov 3 14:42:04 2005 Subject: Transparency in NSImages Message-ID: <199905212224.XAA03101@five-bells.plsys.co.uk> > I have a program in which I want to draw several images on top of one > another, with some of the images having transparent sections so that each > new image doesn't completely overwrite what was drawn before, just the > non-transparent sections. I first create a new, blank image, and then draw > each of the other images on top of it using lockFocus: and > compositeToPoint:. But, after trying many different things, I haven't been > able to get it to work (the transparent part draws white over the previous > image(s)). I created the images using TIFFany, and it assures me that there > is indeed a transparent portion of the image. hasAlpha: even returns YES, > At the risk of being patronising, just because an image has trasparency doesn't mean it is transparent. I presume you have checked to ensure that the sections you think are transparent are indeed see-through? After that, what compositing operator are you using -- NSCompositeSourceOver? You might care to look at the CompositeLab example in /System/Developer/Examples/AppKit/CompositeLab mmalc. From sanguish at digifix.com Fri May 21 15:35:30 1999 From: sanguish at digifix.com (Scott Anguish) Date: Thu Nov 3 14:42:04 2005 Subject: Converting to PDF In-Reply-To: <199905212130.QAA07958@marsh.owlnet.rice.edu> References: <199905212130.QAA07958@marsh.owlnet.rice.edu> Message-ID: <199905212235.SAA21992@digifix.com> Brian Webster wrote: > Mark Tarbell writes: > > > Is there any way to use OmniPDF (or other app) to convert RTF or > > HTML into PDF? What are people using to generate PDF documents? > > There is a version of Ghostscript for the Mac that will convert Postscript > to PDF at > You'd have to use Blue Box and you'd have to save the HTML or RTF to > Postscript, but it'd work eventually. ;-) > FastIO systems has a ClibPDF that will convert TIFF to PDF.. and I imagine that the software could be modified to support HTML and or RTF.. From kcd at jumpgate.com Fri May 21 15:39:54 1999 From: kcd at jumpgate.com (Kenneth C. Dyke) Date: Thu Nov 3 14:42:04 2005 Subject: sgi's XFS + Darwin's VFS? In-Reply-To: <990521151242.399AADoJ.tpugh@rarebear> References: <990521151242.399AADoJ.tpugh@rarebear> Message-ID: <199905212239.PAA16676@babylon5.jumpgate.com> A non-text attachment was scrubbed... Name: not available Type: text/enriched Size: 755 bytes Desc: not available Url : /mailman/archive/macosx-dev/attachments/19990521/22d60a12/attachment.bin From Malcolm_Crawford at plsys.co.uk Fri May 21 17:15:14 1999 From: Malcolm_Crawford at plsys.co.uk (mmalcolm crawford) Date: Thu Nov 3 14:42:04 2005 Subject: AppIcon v0.9 Message-ID: <199905220014.BAA03616@five-bells.plsys.co.uk> Some of you may have noticed that applications use different-sized icons in different situations (e.g. 48x48 in Workspace, but 16x16 in the file menu). You may also have spotted that IconBuilder.app allows you to create and combine images of these different sizes... but sadly the image editing facilities it gives may leave some folks wanting more. So if you're using an alternative editor which has neat graphics features but which doesn't give you a means of combining images, AppIcon might be for you: http://www.plsys.co.uk/~malcolm/AppIcon.app.tar It's not really very sophisticated at the moment -- ideally it should allow multiple image-depths, and do sensible checking -- but hopefully it might prove useful to someone. Best wishes, mmalc. From holger at wizards.de Fri May 21 19:32:09 1999 From: holger at wizards.de (Holger Hoffstaette) Date: Thu Nov 3 14:42:04 2005 Subject: Converting to PDF In-Reply-To: <199905212130.QAA07958@marsh.owlnet.rice.edu> References: <199905212130.QAA07958@marsh.owlnet.rice.edu> Message-ID: <199905220232.EAA11587@ragnarok.en.uunet.de> >Mark Tarbell writes: > >> Is there any way to use OmniPDF (or other app) to convert RTF or >> HTML into PDF? What are people using to generate PDF documents? > >There is a version of Ghostscript for the Mac that will convert Postscript >to PDF at ftp://ftp.cs.wisc.edu/pub/ghost/aladdin/gs550/mac/ The Ghostscript->PDF route is, frankly, crap. What Mark really wants is to evaluate PStill which can be found at http://www.this.net/~frank/pstill.html ..and then send Frank the incredibly small amount of money he asks for. Holger From bwebster at owlnet.rice.edu Fri May 21 21:12:35 1999 From: bwebster at owlnet.rice.edu (Brian Webster) Date: Thu Nov 3 14:42:04 2005 Subject: Transparency in NSImages Message-ID: <199905220423.XAA08007@marsh.owlnet.rice.edu> Thanks everyone, problem solved. I was using NSCompositingCopy mode instead of NSCompositingSourceOver mode. D'oh! -- Brian Webster bwebster@owlnet.rice.edu http://www.owlnet.rice.edu/~bwebster/ From sanguish at digifix.com Sat May 22 00:48:38 1999 From: sanguish at digifix.com (Scott Anguish) Date: Thu Nov 3 14:42:04 2005 Subject: NSDocument/NSTextView crasher Message-ID: <199905220745.DAA24087@digifix.com> I've found a consistent crashing behaviour when using an NSTextView within an NSDocument.. I thought it was just my big app, and then I noticed it in another app I wrote (and they share some code), so I distilled it down to the most basic of basic samples, and it is still dying.. The error is a freedObject error May 22 03:38:43 g3 /Local/Users/sanguish/CrashTest/CrashTest.app/CrashTest: objc: FREED(id): message conversationIdentifier sent to freed object=0x608c8 It seems to be being sent to the NSTextView after the window is closed (in fact, after the entire document has been released)... Anyone up for having a look? I've dropped the crasher app at http://www.digifix.com/CrashTest.tgz (a whooping 11K) To recreate the crash - compile the app - launch it - type something in the text view - hit close - hit "don't save" Kaboom... -- Scott Anguish Middlebury College - Center for Educational Technology Stepwise Server - Mac OS X Server Information From brianw at sounds.WA.COM Sat May 22 05:24:27 1999 From: brianw at sounds.WA.COM (Brian Willoughby) Date: Thu Nov 3 14:42:04 2005 Subject: sgi's XFS + Darwin's VFS? In-Reply-To: <199905212239.PAA16676@babylon5.jumpgate.com> References: <199905212239.PAA16676@babylon5.jumpgate.com> Message-ID: <199905221224.FAA21221@sounds.WA.COM> tpugh@oce.orst.edu wrote: > I would like to second a port of sgi's XFS filesystem if its true. kcd@jumpgate.com replied: [ Actually, something I'd like to see in MacOSX/Darwin before we [ start adding gobs of filesystems to the kernel, and that would [ be... [ [ Dynamically loadable filesystems from some special place on a HD. filesystems aren't (all) added to the kernel, they can be dynamically loaded from /usr/filesystems/*.fs - (I don't know if UFS is part of the kernel). [ This comes from my Amiga days where we had a nice standard for [ doing hard disk partition layout, and a standard way to place [ filesystem code out on the drive. This was great because you [ could unplug your drive and plug it into another machine that [ didn't even have that filesystem and you could still load it. [ It just worked(TM). If your idea is implemented, we will need some sort of security to prevent "guests" from corrupting filesystems by loading their own Trojan fs from a floppy. Darwin could probably be modified to search for loadable filesystems on freshly mounted volumes, but I would hope that a filesystem loaded in this fashion could be restricted to operating only on the volume it was loaded from and no other. On a related note, one of these days I'm going to have to implement a filesystem for the ENSONIQ line of samplers and synthesizers. I have an old NEXTSTEP app which can read the volume as a raw SCSI device so long as it isn't mounted already, but it would be much cleaner if it were a proper filesystem (plus, ATAPI volumes would work, too). NeXT "almost" helped out with this when they heard about my project, but I guess I did not represent "big money" because they declined to provide documentation on how to develop /usr/filesystems. Brian From andrew at stone.com Sat May 22 07:40:44 1999 From: andrew at stone.com (Andrew Stone) Date: Thu Nov 3 14:42:04 2005 Subject: Transparency in NSImages Message-ID: <199905221444.IAA07937@floyd.stone.com> Might as well start writing window system independent code that will also work in OS X user by using: [[NSColor clearColor] set]; NSRectFill(NSMakeRect(0.0,0.0, [image size].width, [image size].height)); instead of : > PSsetgray(NSWhite); > PSsetalpha(0.0); > NSRectFill([image size]); > From: Maury Markowitz .... > The problem is that you have to fill the image with "clear paint" > first. Here, try this just after you create the image... > > [image lockFocus]; > PSsetgray(NSWhite); > PSsetalpha(0.0); > NSRectFill([image size]); > PSsetalpha(1.0); > === Andrew Stone, QuixMaster of Stone Design Corp mailto:andrew@stone.com (505) 345-4800 http://www.stone.com - Create(TM) From andrew at stone.com Sat May 22 08:16:09 1999 From: andrew at stone.com (Andrew Stone) Date: Thu Nov 3 14:42:04 2005 Subject: NSDocument/NSTextView crasher Message-ID: <199905221519.JAA07991@floyd.stone.com> Yep - not much code to be crashing so! Here's the line that causes your grief: [textView setAllowsUndo:YES]; Remove it, and it doesn't crash! Here's your ultra nasty **FIX** for now: In Document.h: - (void)close { [[textView layoutManager]retain]; [super close]; } Mike might have an idea on this - do you have the TextExtras installed? Also, enabling the checkbox "Allows Undo" in the TextView inspector doesn't seem to actually work and the doc edited status doesn't change. Hmmm.... === Andrew Stone, QuixMaster of Stone Design Corp mailto:andrew@stone.com (505) 345-4800 http://www.stone.com - Create(TM) From sanguish at digifix.com Sat May 22 10:17:14 1999 From: sanguish at digifix.com (Scott Anguish) Date: Thu Nov 3 14:42:04 2005 Subject: NSDocument/NSTextView crasher Message-ID: <199905221715.NAA25101@digifix.com> A non-text attachment was scrubbed... Name: not available Type: text/enriched Size: 902 bytes Desc: not available Url : /mailman/archive/macosx-dev/attachments/19990522/81cdab5b/attachment.bin From ash at sineware.com Sat May 22 18:58:58 1999 From: ash at sineware.com (Mishra, Ashish) Date: Thu Nov 3 14:42:04 2005 Subject: MacOSXAmp 0.1 Message-ID: <199905230243.WAA21904@engmail.uwaterloo.ca> > Re: MacOSXAmp 0.1 I got this to work once I checked Preferences, and enable the mpg123 plugin under the Audio IO tab. Once you enable it, choose the Soundkit Output Driver 0.1 Output plugin. And then everything is fine... (mp2s and playing files from your Mac OS partition will also work great :) ash ps. thanks Scott Bender for making life on OS X server that much more fun :) ~~ Ashish Mishra, Sine of the Times, http://www.sineware.com ~~ mac2400, http://www.sineware.com/mac2400 From rcfa at cubiculum.com Sun May 23 01:24:26 1999 From: rcfa at cubiculum.com (Ronald C.F. Antony) Date: Thu Nov 3 14:42:04 2005 Subject: sgi's XFS + Darwin's VFS? In-Reply-To: <990521151242.399AADoJ.tpugh@rarebear> References: <990521151242.399AADoJ.tpugh@rarebear> Message-ID: <199905230824.EAA08302@kannix.cubiculum.com> you wrote: > One last comment before moving this to macosx-talk. > > I would like to second a port of sgi's XFS filesystem if its true. > > Here is a white paper on the XFS filesystem: > http://www.sgi.com/Technology/xfs-whitepaper.html > And a teaser: > > 1.1 Product Feature Summary Just a word of warning: SGI said that not all XFS features will be made open/free. That means XLV, guaranteed rate I/O and some of the other non-basic features may NOT be part of what they release. But even a basic 64-bit journalling file system is already a good step in the right direction, and with fast disks and a hardware RAID, many of the advanced features are not required in the typical Mac multi-media environment. Ronald ============================================================================== "The reasonable man adapts himself to the world; the unreasonable one persists in trying to adapt the world to himself. Therefore all progress depends on the unreasonable man." G.B. Shaw | rcfa@cubiculum.com | NeXT-mail welcome From rcfa at cubiculum.com Sun May 23 05:01:58 1999 From: rcfa at cubiculum.com (Ronald C.F. Antony) Date: Thu Nov 3 14:42:04 2005 Subject: sgi's XFS + Darwin's VFS? In-Reply-To: <199905221224.FAA21221@sounds.WA.COM> References: <199905221224.FAA21221@sounds.WA.COM> Message-ID: <199905231202.IAA09078@kannix.cubiculum.com> > filesystems aren't (all) added to the kernel, they can be dynamically loaded > from /usr/filesystems/*.fs - (I don't know if UFS is part of the kernel). Of course, those types one might want to use as boot/root file system, should probably be in the kernel. Loadable filesystems are best used for network file systems, file system abstractions and foreign file systems generally used on non-primary partitions and removable media. In regards to security: it's probably best if only file systems installed on the root partition and owned by root are eligible for loading into the kernel. New file systems types that could be loaded from removable media would best be implemented as a user-level NFS server, since that would isolate the code sufficiently from the kernel. Also, restricting the code to the removable disk that contains the code and/or to file systems manually mounted by root, would be a good idea, in terms of security. Ronald ============================================================================== "The reasonable man adapts himself to the world; the unreasonable one persists in trying to adapt the world to himself. Therefore all progress depends on the unreasonable man." G.B. Shaw | rcfa@cubiculum.com | NeXT-mail welcome From andrew at stone.com Sun May 23 08:58:37 1999 From: andrew at stone.com (Andrew Stone) Date: Thu Nov 3 14:42:04 2005 Subject: ANNOUNCE: Stone Design Announces NEW LOW prices Message-ID: <199905231601.KAA10648@floyd.stone.com> (I apologize for crossposting - but this is good news, especially for students!) **** NEWS For Immediate Release For more information, please contact: Katie Graunke TEL: 505-345-4800 FAX: 505-345-3424 http://www.stone.com/pr/Low_Low_Prices.html STONE DESIGN Announces Special Intro Pricing For Create and Web Tools Albuquerque, May 24 1999 - Veteran Macintosh and Cocoa developers, Stone Design, today announced special introductory pricing for Create(tm) and the Stone Design Web Tools Suite for Mac OS X Server and WebObjects 4.0. For a limited time, Create is available for just $199 (list: $695) and the Stone Design Web Tools Suite is just $ 299 (list: $864). For bona fide students who fax in a current ID, Create is $ 99, and the Suite is $149. Create 5.0 is the premier multipage draw studio with instant graphic conversion, web page authoring, autotracing, PostScript programming environment, patterns, layers, rulers, neon, special effects, Super Text and much, much more! Stone Design Web Tools Suite contains Create, GIFfun, Slice And Dice, Pack Up And Go, and WebColors. HURRY! These prices won't last. Downloading is available now - please visit www.stone.com today. +=+ Stone Design Corp of Albuquerque New Mexico was formed in 1984 and was the first developer to ship a product on NeXT Hardware in 1989 with the release of TextArt. Since then, Stone Design has been a leader in innovative shrinkwrap software in the NEXTSTEP/OpenStep/Mac OS X Server marketplace. Create(TM) - the high end drawing and web page making package - is Mac OS X Server's and Apple Yellowbox's first application. All other names are trademarked by their respective companies. === Andrew Stone, QuixMaster of Stone Design Corp mailto:andrew@stone.com (505) 345-4800 http://www.stone.com - Create(TM) From holger at wizards.de Sun May 23 12:00:06 1999 From: holger at wizards.de (Holger Hoffstaette) Date: Thu Nov 3 14:42:04 2005 Subject: 4.2 DriverKit questions / help needed Message-ID: <199905231900.VAA02671@ragnarok.en.uunet.de> Hi, Some of you might have seen my posting to csn.programmer recently. I'm looking for a way to make the monitor go into power saving mode from user mode - either by invoking some APM related function or just pulling one of the SYNC signals so that the monitor shuts off by itself. So far, I have managed to correctly ask my card for its RAMDAC speed via IODeviceMaster and it's generic get/set facility. Unfortunately that's about the only useful thing I managed to do: except for getting the current display mode number, everything else returns with errors (not implemented, invalid argument, or -750 (undefined)). Other problems: - even something relatively simple as getting the IO_GET_DISPLAY_MODE_INFO struct fails every time, no matter what I use as parameter count. - the IODevice class method for looking up a driver instance seems to be unavailable in user mode, i.e. from a simple app. Is this correct? I've tried to look up the id pointer via the driver's object number, but only get back nil. I also had a look at the Darwin DriverKit, but it seems some things have been added, e.g. IOPower.h, user-level Mach port lookup and more #defines for display mode information (width, height etc.), which makes developing with 4.2 'harder than necessary'. :-( If anybody with DriverKit experience would be so kind and offer to look at my little example code I would be very thankful. It's very small, but I have the nagging feeling that I'm missing something trivial. Also, does anybody know if the source to Scott Vail's ModeControlLoginUI.loginbundle is available? Is Scott still at Apple? If so - Scott, if you happen to read this, please be so kind to drop me a line. Thanks, Holger From karmat at matthias.org Sun May 23 14:46:23 1999 From: karmat at matthias.org (Karl N. Matthias) Date: Thu Nov 3 14:42:04 2005 Subject: ANNOUNCE: MacOSXAmp 0.1 is available In-Reply-To: <9905191753.AA061505@bass> Message-ID: Some people feel that standard user interfaces are not necessarily well-suited to all applications. If you don't like it, don't use those applications. No one is making you. But please, people, lets try not to flame other developers on here. I'm clearly not the moderator here, but I feel like people tend to reply to things without thinking of how they might sound to others, especially the person writing the piece of software in question. Thanks Karl On Wed, 19 May 1999, Christian Pekeler wrote: > [MacOSXAmp] > > The main reason for this release is to get feedback > > Ok. After looking at http://www.harmony-ds.com/MacOSXAmp/shot1.jpg and http://www.harmony-ds.com/MacOSXAmp/shot2.jpg I can't stop myself from ranting. This application like so many other sound-related applications ignores all GUI standards. I was a great fan of NEXTSTEP because nearly all applications look similar and consistent. This is a great advantage everyone should understand. As far as I know the MacOS it has the same goal and that's what the Mac is famous for. There is no conceptional difference between a CDPlayer for instance and other applications like mailviewer which justifies that the user has to learn yet another GUI. If the programmer wants to show her creativity, then do a fancy info panel. But please don't experiment with the main gui. > > If I set that the main backgroundcolor of windows should be blue instead of grey in the system's preferences, I want ALL applications to apply to my settings. If I change the system font to a bigger size because I'm visually impaired, I want ALL applications to apply to my settings. If Apple updates the AppKit to change the appearance of buttons, I want all applications to have this new type of button. Will that be possible with MacOSXAmp's custom gui? > > I don't care if Windows or X-windows applications invent their own guis. But please don't do this on my favorite platform. I think every application developer has his share of responsibility how the public perceives the operating system as a whole. > > Please take a look at Apple/Documentation/Developer/YellowBox/ReleaseNotes/AppLayout.html for a start. > > Sorry, but I had to let this out. > Christian > > From jcr at idiom.com Sun May 23 19:59:55 1999 From: jcr at idiom.com (John C. Randolph) Date: Thu Nov 3 14:42:04 2005 Subject: NSDocument class: How do I dynamically set the extensions I'll take? Message-ID: <3748C09D.CD875530@idiom.com> Gentlemen, I'm writing a little image viewing app for practice, and I'd like to be able to open any and all of the image file types that NSImage will allow. I know how to get the list of those types, but I haven't figured out how to make my NSDocument and NSDocumentController adjust dynamically. Any thoughts? Thanks, -jcr From gabriel at accentcomm.com Mon May 24 03:32:45 1999 From: gabriel at accentcomm.com (Gabriel Androczky) Date: Thu Nov 3 14:42:04 2005 Subject: ANNOUNCE: MacOSXAmp 0.1 is available References: Message-ID: <37492ACD.1DE0EB88@accentcomm.com> "Karl N. Matthias" wrote: > > Some people feel that standard user interfaces are not necessarily > well-suited to all applications. If you don't like it, don't use > those applications. No one is making you. But please, people, lets > try not to flame other developers on here. I'm clearly not the > moderator here, but I feel like people tend to reply to things without >thinking of how they might sound to others, especially the person > writing the piece of software in question. > Sure, But _someone_ has to tell his/her feelings to the developer. If everyoune would go and develop some own GUI, the whole OS would quickly become a whole mess, and that is not what Macintosh (and the old NeXT way) is about. A consistent GUI gives at all times the same feeling and contains the same kind of elements that most of the users have already seen or used. Don't get me wrong, I like WinAmp and MacOSXAmp, but since it needs something "not standard" for a GUI, why not adopting something that is also new and will probably get standard in the Apple world? For example, the way Quicktime 4 looks. I can imagine something similar but not as a separate skin or so since most people will _use_ it and not decorate it. Regards, Gabriel -- -------------------------------------------------------------- A C C E N T C O M M U N I C A T I O N S - We Make Your Web Groove - Tel: (+36 1) 315-0701 http://www.accentcomm.com GSM: (+36 30) 951-4334 e-mail: info@accentcomm.com Fax: (+36 30) 980-0344 -------------------------------------------------------------- From andrew at stone.com Mon May 24 05:26:46 1999 From: andrew at stone.com (Andrew Stone) Date: Thu Nov 3 14:42:04 2005 Subject: NSDocument class: How do I dynamically set the extensions I'll take? Message-ID: <199905241227.GAA12726@floyd.stone.com> > Gentlemen, Are there no Gentlewomen amongst us? Alas what a geeky profession! > I'm writing a little image viewing app for practice, and I'd like to be > able to open any and all of the image file types that NSImage will > allow. I know how to get the list of those types, but I haven't figured > out how to make my NSDocument and NSDocumentController adjust dynamically. Some of the Cocoa engineers agreed with me that this is a hole in the NSDocument/NSWorkspace/PB file extension naming mechanism - that is - there should be a switch on the PB Project Attributes inspector which includes "Image types", including the dynamically available types available through filter services at runtime. But there isn't, so viva la forca bruta: a. In gdb, print the contents of NSImage fileTypes: (gdb) po [NSImage imageFileTypes] ( tif, tiff, PNG, png, GIF, gif, JPG, jpg, JPEG, jpeg, CUR, cur, ICO, ico, BMP, bmp, TIF, TIFF, eps, PICT, pict ) b. For each type, assign an icon. c. Again, note this won't handle types that are dynamically added to NSImage === Andrew Stone, QuixMaster of Stone Design Corp mailto:andrew@stone.com (505) 345-4800 http://www.stone.com - Create(TM) From paulrs at lgs-systems.com Mon May 24 08:07:17 1999 From: paulrs at lgs-systems.com (Paul R. Summermatter) Date: Thu Nov 3 14:42:04 2005 Subject: NSPopUpButton, centering images & MenuMadness Message-ID: <9905241507.AA00924@slab> Hey guys, I'm trying to replace a custom pop up button that we built which we were using in the corner views of table views (just like the EOF 2.2 version of EOModeler). The MenuMadness example nicely demonstrates the new menu & pop up button API, but there's one problem which I can't seem to workaround, and that is the issue of centering the image in the pop up. I've configured the pop up button cell to have NSPopUpNoArrow, but that doesn't seem to make a difference. As an example of what I am experiencing, take the MenuMadness example, and modify either the noSelPullDown or noSelPopUp to have NSPopUpNoArrow and a frame with a width of 20.0 and a height of 20.0 (just a little bigger than the image and just about the size of a corner view in a table view). Unlike a normal button with an imagePosition of NSImageOnly where the image is centered, the pop up button keeps pushing the image over, making it off center. It seems as if there is a mandatory left padding on NSMenuItems and NSMenuItemCells, even when one doesn't really make sense in a button like this. Does anyone know of a workaround to this? Is there another way to implement pulldown menus in the corner views of tables using only appkit classes? Any thoughts are welcome. Regards, Paul --- Paul Summermatter LGS Systems, Inc. Medical Computing Division 15 TJ Gamester Ave Portsmouth, NH 03801-5871 (603) 433-9822 voice (603) 433-9818 fax (888) 898-6321 pager 8986321@skytel.com paging email paulrs@lgs-systems.com (NeXT or MIME Mail Welcome) http://www.lgs-systems.com From jjfeiler at relief.com Mon May 24 09:46:57 1999 From: jjfeiler at relief.com (John Jay Feiler) Date: Thu Nov 3 14:42:04 2005 Subject: NSColorWells in an NSTableView Message-ID: <199905232044.NAA25427@haque.relief.com> Anybody have any good suggestions for getting an NSColorWell into an NSTableView? Most other widgets work great, since thery're based around cells, but NSColorWell doesn't have a corresponding NSColorWellCell. J -- John Feiler jjfeiler@relief.com From theisen at akaMail.com Mon May 24 12:03:05 1999 From: theisen at akaMail.com (Dirk Theisen) Date: Thu Nov 3 14:42:04 2005 Subject: Subclassing NSSplitView? Bug? Message-ID: <1dsbr3q.17p4v0gs7ak1wM@ascend-tk-p87.rhrz.uni-bonn.de> Hello! Is there any special trick involved subclassing NSSplitView (in Java)? I tried this with DR2, but but thought this might have been unimplemented. Now (running MXS) I think this might be a bug: The Java tutorial shows how to subclass NSView in Java. I did the same with NSSplitView, but whenever I try to run my app containing the nib with the NSSplitView subclass instance defined I get an NSArchiverArchiveInconsistencyException (I have to boot into MacOS to send this (being on PPP) so the exact word might be wrong). It does not matter if my class defines instance variables or not. If anyone is willing to look at it: I can provide access to a project showing the behaviour. Any tips welcome. Regards, Dirk -- Buy a Pentium III now, get your personal Big Brother FREE!!! http://theisen.home.pages.de/ From josh at i33.com Mon May 24 12:03:13 1999 From: josh at i33.com (Josh Flowers) Date: Thu Nov 3 14:42:04 2005 Subject: Jasmine? Message-ID: <199905241903.PAA17897@fnord.i33.com> While looking into possible object-oriented database solutions for my current project, I came across an old press release about Computer Associates Jasmine product being integrated with Yellow Box and Rhapsody. Does anyone know what's happened (or is happening) with this? From paulrs at lgs-systems.com Mon May 24 11:31:17 1999 From: paulrs at lgs-systems.com (Paul R. Summermatter) Date: Thu Nov 3 14:42:04 2005 Subject: NSColorWells in an NSTableView In-Reply-To: <199905232044.NAA25427@haque.relief.com> References: <199905232044.NAA25427@haque.relief.com> Message-ID: <9905241831.AA01093@slab> You wrote: [ Anybody have any good suggestions for getting an NSColorWell into an [ NSTableView? Most other widgets work great, since thery're based [ around cells, but NSColorWell doesn't have a corresponding [ NSColorWellCell. [ [ J [ [ [ -- [ John Feiler [ jjfeiler@relief.com [ John et al, This is annoying. However, there may have been a good reason for it. At first, I thought maybe the cell was just hidden, but, at least on OS 4.2 Mach, I can find no special color cell class in the AppKit class dump. In any event, it looks to me like some work has been done on the NSColorPanel, so maybe one of the engineers will open up or fix this API for the OS X release. Regards, Paul --- Paul Summermatter LGS Systems, Inc. Medical Computing Division 15 TJ Gamester Ave Portsmouth, NH 03801-5871 (603) 433-9822 voice (603) 433-9818 fax (888) 898-6321 pager 8986321@skytel.com paging email paulrs@lgs-systems.com (NeXT or MIME Mail Welcome) http://www.lgs-systems.com From jhendry at cmg.FCNBD.COM Mon May 24 13:38:31 1999 From: jhendry at cmg.FCNBD.COM (Jonathan Hendry) Date: Thu Nov 3 14:42:04 2005 Subject: NSColorWells in an NSTableView In-Reply-To: <9905241831.AA01093@slab> References: <9905241831.AA01093@slab> Message-ID: <199905242038.PAA15412@wo1085.cmg.FCNBD.COM> > You wrote: > [ Anybody have any good suggestions for getting an NSColorWell into an > [ NSTableView? Most other widgets work great, since thery're based > [ around cells, but NSColorWell doesn't have a corresponding > [ NSColorWellCell. > > John et al, > > This is annoying. However, there may have been a good reason for > it. At first, I thought maybe the cell was just hidden, but, at least on OS > 4.2 Mach, I can find no special color cell class in the AppKit class dump. > In any event, it looks to me like some work has been done on the > NSColorPanel, so maybe one of the engineers will open up or fix this > API for the OS X release. How much of NSColorWell's functionality is needed? An NSColor will draw the 'swatch' if you call - (void)drawSwatchInRect:(NSRect)rect; All that's really needed to duplicate an NSColorWell is a cell that uses NSColors to draw a swatch in itself, which accepts colors dropped on it, which can bring up the color panel when clicked, and which can enter a mode in which it listens for NSColorPanelColorDidChangeNotification and sets its color accordingly. (Of course, it *would* be easier if there were an NSColorWellCell.) - Jon From karl at nfox.com Mon May 24 15:55:53 1999 From: karl at nfox.com (Karl Kraft) Date: Thu Nov 3 14:42:04 2005 Subject: NSColorWells in an NSTableView Message-ID: <003f01bea638$96785340$0328040a@kiva.ensuing.net> Same problem here. I played with many ideas before I stumbled across a cheap hack. Put a NSColorWell outside the visible area of the window. Write your own cell to draw the swatch into the table view. On a mouse down on your cell, select the hidden NSColorWell, and set the target to a method of your cell. The downside, is that things like shift selecting to get multiple color wells selected at once become difficult. (you could do it by creating multiple wells off-window.) In our case, it didn't affect our UI to a great deal, so we didn't go through the trouble. -----Original Message----- From: John Jay Feiler To: Multiple recipients of list Date: Monday, May 24, 1999 9:48 AM Subject: NSColorWells in an NSTableView Anybody have any good suggestions for getting an NSColorWell into an NSTableView? Most other widgets work great, since thery're based around cells, but NSColorWell doesn't have a corresponding NSColorWellCell. J -- John Feiler jjfeiler@relief.com From epeyton at epicware.com Mon May 24 12:34:28 1999 From: epeyton at epicware.com (Eric Peyton) Date: Thu Nov 3 14:42:04 2005 Subject: OpenGL development on X Server Message-ID: All, First off if you haven't played with the new Omnigroup Accelerated MesaGL/Glide implementation I would recommend doing it. Mac OS X Server is now closer to becoming a kick ass game/graphics development platform for Apple and the world. Thanks Omni! With Time Wood's help I have completed porting more than 20 screen savers/examples to Omni's mesagl (If you already have a voodoo2 card installed and want to play with it drop me a line and I will forward you the savers). I will be releasing the savers open source "Real Soon Now"(tm). What I am looking for from our developer community is suggestions on MesaGL/OpenGL projects that should/could be ported to Mac OS X Server. I am willing to work on some stuff like this and I hope others are as well and if you know of a project and/or source that would be nice to have, please let me know. Thanks, Eric -- Eric Peyton epeyton@epicware.com Software and Source for Mac OS X Server From epeyton at apple.com Mon May 24 19:34:08 1999 From: epeyton at apple.com (Eric Peyton) Date: Thu Nov 3 14:42:04 2005 Subject: OpenGL development on X Server Message-ID: <199905250232.TAA28500@scv3.apple.com> There has always been a version of Mesa which is not accelerated. 4 or 5 different versions at different times if I recall. I have one version available for download (binary only, let me know if you want the source) at http://www.epicware.com/glsavers.html. I can't remember who did this port. I have used this version for about a year now, so I am sure it's out of date. Making Mesa into a framework is very easy ... Eric > From: Yves Starreveld > Date: 1999-05-24 22:22:00 -0400 > To: epeyton@epicware.com > Subject: Re: OpenGL development on X Server > Delivered-To: epeyton@enteract.com > X-Accept-Language: en > X-Mailer: Mozilla 4.6 (Macintosh; I; PPC) > > Hi, Eric, > I don't have a voodoo2 card, but is there a version of Mesa which will > run in software on OSX? > I would nominate vtk, the Visualization toolkit (www.kitware.com) as a > good candidate. I am actually working on it a lot, and will be porting > it myself. Downsides are the current need for an X-server. vtk makes an > amazing visualization platform, and its extensibility, flexibility and > availability of source code are all big pluses. > > Yves > > Yves Starreveld, MD > Robarts Research Lab > Image Guided Surgery Group > -- Eric Peyton Senior Consulting Engineer Apple Computer, Inc. epeyton@apple.com From erik at object-factory.com Tue May 25 03:33:56 1999 From: erik at object-factory.com (Erik Doernenburg) Date: Thu Nov 3 14:42:04 2005 Subject: NSColorWells in an NSTableView Message-ID: <199905251035.DAA27479@ignem.omnigroup.com> > You wrote: > [ Anybody have any good suggestions for getting an NSColorWell into an > [ NSTableView? Most other widgets work great, since thery're based > [ around cells, but NSColorWell doesn't have a corresponding > [ NSColorWellCell. > > John et al, > > This is annoying. However, there may have been a good reason for > it. At first, I thought maybe the cell was just hidden, but, at least on OS > 4.2 Mach, I can find no special color cell class in the AppKit class dump. > In any event, it looks to me like some work has been done on the > NSColorPanel, so maybe one of the engineers will open up or fix this API for > the OS X release. Remember that color wells rely on drag & drop and cells cannot register themselves as dragging destination. This makes a cell based color well rather unattractive. If you don't need drop support you can listen for NSColorPanelColorDidChangeNotifications and draw a little coloured area yourself. regards, erik -- OBJECT FACTORY, Gesellschaft fuer Informatik und Datenverarbeitung mbH Telephone: ++49 +231 9751370 . Internet: http://www.object-factory.com From maury at OAAI.COM Tue May 25 08:10:28 1999 From: maury at OAAI.COM (Maury Markowitz) Date: Thu Nov 3 14:42:04 2005 Subject: NSDocument class: How do I dynamically set the extensions I'll take? Message-ID: <199905251506.LAA00658@OAAI.COM> > I'm writing a little image viewing app for practice, and I'd like to be > able to open any and all of the image file types that NSImage will > allow. I know how to get the list of those types, but I haven't figured > out how to make my NSDocument and NSDocumentController adjust dynamically. As far as I was able to tell, you can't get this to work. I tried overriding writableTypes and providing my own list, but it doesn't work. I ended up hard coding into the plist (sigh). While on the topic, I think this is one problem in Cocoa now - the profusion of types. I think we should have a single NSType object which maintains all of this information for us, and is passed into the various NSPasteboard methods instead of the string we use currently. Maury From Steve.Ivy at cbn.org Tue May 25 08:14:54 1999 From: Steve.Ivy at cbn.org (Ivy, Steven) Date: Thu Nov 3 14:42:04 2005 Subject: plist format Message-ID: <2FAD21AE9E37D21191F900805FE67787322054@NT2> Can anyone out there point me to a quick tutorial on plists? I'm working on a little test app which uses NSarrays and NSDicts, and I want to be able to use writeToFile: and initFromFile:... Anyway, I'm trying to understand how each are written out, and I want to write out a quick and dirty default plist to read from to initialize the object. I've looked through a bunch of .plists on my OSXS box, but haven't quite deduced the formatting. Help? --Steve ---------------------------------------------------- "Should we talk about the weather? Should we talk about the government?" -REM Steve Ivy, steve.ivy@cbn.org ---------------------------------------------------- From richard at brainstorm.co.uk Tue May 25 08:38:50 1999 From: richard at brainstorm.co.uk (Richard Frith-Macdonald) Date: Thu Nov 3 14:42:05 2005 Subject: plist format In-Reply-To: <2FAD21AE9E37D21191F900805FE67787322054@NT2> References: <2FAD21AE9E37D21191F900805FE67787322054@NT2> Message-ID: <9905251538.AA00675@tiptree.brainstorm.co.uk> On Tue, 25 May 1999 08:26:31 -0700 (PDT), Steve.Ivy@cbn.org wrote: > Can anyone out there point me to a quick tutorial on plists? I'm > working on a little test app which uses NSarrays and NSDicts, and I want to > be able to use writeToFile: and initFromFile:... Anyway, I'm trying to > understand how each are written out, and I want to write out a quick and > dirty default plist to read from to initialize the object. > > I've looked through a bunch of .plists on my OSXS box, but haven't > quite deduced the formatting. Help? To read a property-list from a text file - myString = [NSString stringWithContentsOfFile: file]; result = [myString propertyList]; To write a property-list to a file - myString = [result description]; myData = [myString dataUsingEncoding: NSASCIIStringEncoding]; [myData writieToFile: file atomically: YES]; >From the help for the GNUstep 'defaults' command - A property list is a method of providing structured information consisting of strings, arrays, dictionaries, and binary data. In a property list, strings appear as plain text (as long as they contain no special characters), and inside quotation marks otherwise. Special characters inside a quoted string are 'escaped' by a backslash. This escape mechanism is used to permit the double quote mark to appear inside a quoted string. Arrays appear as a comma separated list of items delimited by brackets. Dictionaries appear as a series of key-value pairs, each pair is followed by a semicolon and the whole dictionary is delimited by curly brackets. Data is encoded as hexadecimal digits delimited by angle brackets. Here is an example of a dictionary encoded as a text property list - { Name = "My Application"; Author = "Just me and \"my other half\""; Modules = ( Main, "'Input output'", Computation ); Checksum = <01014b5b 123a8b20> } --- `Everything should be made as simple as possible, but not simpler.' From ryan at omnigroup.com Tue May 25 10:19:55 1999 From: ryan at omnigroup.com (Ryan Dingman) Date: Thu Nov 3 14:42:05 2005 Subject: plist format Message-ID: <199905251719.KAA11273@tapeworm.omnigroup.com> > Can anyone out there point me to a quick tutorial on plists? I'm > working on a little test app which uses NSarrays and NSDicts, and I want to > be able to use writeToFile: and initFromFile:... Anyway, I'm trying to > understand how each are written out, and I want to write out a quick and > dirty default plist to read from to initialize the object. > > I've looked through a bunch of .plists on my OSXS box, but haven't > quite deduced the formatting. Help? The documentation for plists is in /System/Documentation/Developer/YellowBox/TasksAndConcepts/ProgrammingTopics/PropertyLists.pdf Hope this helps. ryan -- Ryan Dingman Omni Development, Inc. ryan@omnigroup.com From Steve.Ivy at cbn.org Tue May 25 10:28:09 1999 From: Steve.Ivy at cbn.org (Ivy, Steven) Date: Thu Nov 3 14:42:05 2005 Subject: plist format Message-ID: <2FAD21AE9E37D21191F900805FE6778732205A@NT2> Thanks for all the replies- I found the dolcumentation. --Steve ---------------------------------------------------- "Should we talk about the weather? Should we talk about the government?" -REM Steve Ivy, steve.ivy@cbn.org ---------------------------------------------------- > ---------- > From: Ryan Dingman > Reply To: ryan@omnigroup.com > Sent: Tuesday, May 25, 1999 1:22 PM > To: Multiple recipients of list > Subject: Re: plist format > > > Can anyone out there point me to a quick tutorial on plists? I'm > > working on a little test app which uses NSarrays and NSDicts, and > I want to > > be able to use writeToFile: and initFromFile:... Anyway, I'm trying to > > understand how each are written out, and I want to write out a > quick and > > dirty default plist to read from to initialize the object. > > > > I've looked through a bunch of .plists on my OSXS box, but > haven't > > quite deduced the formatting. Help? > > The documentation for plists is in > > /System/Documentation/Developer/YellowBox/TasksAndConcepts/ProgrammingTopi > cs/PropertyLists.pdf > > > Hope this helps. > > ryan > > -- > Ryan Dingman > Omni Development, Inc. > ryan@omnigroup.com > From theisen at akaMail.com Tue May 25 11:15:17 1999 From: theisen at akaMail.com (Dirk Theisen) Date: Thu Nov 3 14:42:05 2005 Subject: Vertical Splitviews in IB? Message-ID: <1dsdmcc.l1qr6u4mjiipM@ascend-tk-p87.rhrz.uni-bonn.de> Hi! This is probably a FAQ, but anyway: How can I specify an NSSplitView to be vertical (i.e. dividing in left and right view) within IB? I see the method is there, but the inspector for the splitiew in IB is empty. Regards, Dirk [searching for the trick] -- Buy a Pentium III now, get your personal Big Brother FREE!!! http://theisen.home.pages.de/ From gerard.iglesias at ina.fr Tue May 25 13:54:12 1999 From: gerard.iglesias at ina.fr (Gerard Iglesias) Date: Thu Nov 3 14:42:05 2005 Subject: Vertical Splitviews in IB? References: <1dsdmcc.l1qr6u4mjiipM@ascend-tk-p87.rhrz.uni-bonn.de> Message-ID: <374B0DF4.FC5E9FA6@ina.fr> Dirk Theisen wrote: > > How can I specify an NSSplitView to be vertical (i.e. dividing in left > and right view) within IB? Position you subview in the right position: right side of the first near the left side of the second, hence when you will choose "Group in split view", you will obtain what you want. Regards. -- Gerard Iglesias Ph.D. in Computer Graphic gerard.iglesias@ina.fr Researcher/Project Manager http://www.mediaport.net/People/Gerard.IGLESIAS From brianw at irisgroup.com Tue May 25 19:35:04 1999 From: brianw at irisgroup.com (Brian Wagner) Date: Thu Nov 3 14:42:05 2005 Subject: JFIF & GIF, Thanks References: <199905210319.UAA05090@scv3.apple.com> Message-ID: <374B5DD7.DAEEB9D5@irisgroup.com> Thanks for the code snippet. This is I need to learn the language. Does anybody know where to get sample code for Ojective-C. Apple's examples are a little thin. Brian Wagner ======================================================== Eric Peyton wrote: > Here's what I would do in ObjC - you can try to translate it to Java ... > > // init the image rep (2 lines of code) > > NSImage *image; > NSString *myString = @"MyString" > > // Fill the bitmap with the buffer > image = [[NSImage alloc] initWithSize:NSMakeSize(200,200)]; > > [image lockFocus]; > > // draw the string > > [myString drawAtPoint:NSMakePoint(0,0)]; > > [image unlockFocus]; > > // save the image to a file > [[image TIFFRepresentation] writeToFile:@"FilePathGoesHere" > atomically:NO]; > > Of course that is untested - but it should work for you > > (And of course you can use an NSAttributedString instead of an > NSString and get all sorts of fonting and sizing and coloring till > your hearts content ...) > From sanguish at digifix.com Wed May 26 00:25:51 1999 From: sanguish at digifix.com (Scott Anguish) Date: Thu Nov 3 14:42:05 2005 Subject: Is this a multi-threaded problem? Message-ID: <199905260722.DAA11052@digifix.com> I've been tracking down a bug in NicCentral that I was told about today... its an odd one... For those who don't know, NicCentral does basic network lookup type stuff... (http://www.stepwise.com/Software/NicCentral) Mainly what it does is calls BSD stuff to do things like nslookup or what have you.. Because some of these calls can take a while, and you may want to have multiple windows actively doing searches, I've attempted to make the app multi-threaded. I say attempted, because it seems to work great, but an odd problem has cropped up that I think might be related to the multi-threading. If you run the strobe tool and then go to the Info panel... the app dies with a freed object message in the Info panel stuff... I've already looked at the actual info panel code... so I'm now looking into this.. because it doesn't crash until you reach a certain point in the process... This gets called after setting a bunch of stuff in the dictionary (the command, etc..) [NSThread detachNewThreadSelector:@selector(executeCommand:) toTarget:self withObject:theDict]; I am taking the precaution of using an NSLock around the actual insertion of text into the textview (the insertReBoldText: for example goes through this... although it sets up a bunch of stuff first..) But I think this chunk of code is where things are going wrong.... Since this will be the approach I use in OpenUp 3.0 (coming soon) I'd appreciate any help... - (void)executeCommand:(NSDictionary *)executionData { NSTask *pipeTask; NSPipe *readPipe,*errorPipe; NSFileHandle *readHandle,*errorHandle; NSData *inData = nil; NSData *errorData = nil; NSMutableData *outData; NSAutoreleasePool *myAutoreleasePool; myAutoreleasePool=[[NSAutoreleasePool alloc] init]; pipeTask = [[NSTask alloc] init]; readPipe = [NSPipe pipe]; errorPipe = [NSPipe pipe]; readHandle = [readPipe fileHandleForReading]; errorHandle = [errorPipe fileHandleForReading]; outData=[NSMutableData data]; if ([executionData objectForKey:@"directoryPath"]) { [pipeTask setCurrentDirectoryPath:[executionData objectForKey:@"directoryPath"]]; } [pipeTask setStandardOutput:readPipe]; [pipeTask setStandardError:errorPipe]; [pipeTask setLaunchPath:[executionData objectForKey:@"Command"]]; if ([executionData objectForKey:@"Args"]) [pipeTask setArguments:[executionData objectForKey:@"Args"]]; [pipeTask launch]; if ([pipeTask isRunning]) { while (((inData = [readHandle availableData]) && [inData length]) ) { if (stop) { stop=NO; [self insertRedBoldText:@"\n---- Cancelled by user ----\n"]; [self enableStart]; [myAutoreleasePool release]; [NSThread exit]; } if (inData) [self insertPlainText:[[[NSString alloc] initWithData:inData encoding:NSASCIIStringEncoding] autorelease]]; } } [pipeTask waitUntilExit]; if ([pipeTask terminationStatus]) { while ((errorData = [errorHandle availableData]) && [errorData length]) { if (errorData) { [outData appendData:errorData]; } } [self insertRedBoldText:@"\nError Encountered\n"]; [self insertBoldText:[[[NSString alloc] initWithData:outData encoding:NSASCIIStringEncoding] autorelease]]; [self insertRedBoldText:@"\n--------\n\n\n"]; } else [self insertBoldText:@"\n---- Completed ----\n\n\n"]; [pipeTask release]; [self enableStart]; [myAutoreleasePool release]; stop=YES; return; } -- Scott Anguish Middlebury College - Center for Educational Technology Stepwise Server - Mac OS X Server Information From sanguish at digifix.com Wed May 26 00:54:56 1999 From: sanguish at digifix.com (Scott Anguish) Date: Thu Nov 3 14:42:05 2005 Subject: On the topic of detaching threads.. Message-ID: <199905260752.DAA11096@digifix.com> The problem that I was encountering seems to be fixed by moving to an NSWindowController and the notification hack that I was sent in response to the NSDocument crasher I was asking about the other day.. But the question about how to handle multi-threading still stands.... The current problem that I'm finding is that if a user cancels the action mid stream, that it can take a substantial time to return because the thread is still executing.. if the document gets closed before the thread finishes (cancelled or not) the stuff in the detached thread that puts the information into the textview still gets called, and that causes a crash.. So my question (long winded I know).. How can I immediately stop the detached thread? Is that possible? That way if the user clicks the close box, I can just kill the thread and let it happen.. likewise, if they cancel the action (like a strobe that can take a while) I can immediately return.. Any help would be much appreciated... -- Scott Anguish Middlebury College - Center for Educational Technology Stepwise Server - Mac OS X Server Information From sanguish at digifix.com Wed May 26 01:02:01 1999 From: sanguish at digifix.com (Scott Anguish) Date: Thu Nov 3 14:42:05 2005 Subject: NSDocument/NSTextView crasher Message-ID: <199905260802.EAA11151@digifix.com> I noticed that this wasn't posted by Mike to the list.. so I figured I'd forward it (since it is important).. Andrew's fix caused a leak... this doesn't.. Thanks again to Mike and Andrew (and everyone else who sent comments) Mike said: This is a known problem that has been fixed in OSX DP1 (it doesn't happen to you there, does it?) A safe workaround that will not cause leaks would be to add the following to your Document class in the test case: - (void)windowWillClose:(NSNotification *)notification { [[notification object] makeFirstResponder:[notification object]]; } The problem is that the NSInputManager is not noticing that the NSTextView goes away, so getting the text view to give up first responder makes it notice. This will work before and after the fix. From erik at object-factory.com Wed May 26 02:06:31 1999 From: erik at object-factory.com (Erik Doernenburg) Date: Thu Nov 3 14:42:05 2005 Subject: Is this a multi-threaded problem? Message-ID: <199905260906.CAA25171@ignem.omnigroup.com> > Mainly what it does is calls BSD stuff to do things like nslookup or > what have you.. > > Because some of these calls can take a while, and you may want to > have multiple windows actively doing searches, I've attempted to make > the app multi-threaded. I say attempted, because it seems to work > great, but an odd problem has cropped up that I think might be > related to the multi-threading. > > [...] > > I am taking the precaution of using an NSLock around the actual > insertion of text into the textview (the insertReBoldText: for > example goes through this... although it sets up a bunch of stuff > first..) I don't think a lock is enough because you might still interupt layout managers that work in the background. The only really safe solution is to go through an NSConnection. > The current problem that I'm finding is that if a user cancels the > action mid stream, that it can take a substantial time to return > because the thread is still executing.. if the document gets closed > before the thread finishes (cancelled or not) the stuff in the > detached thread that puts the information into the textview still > gets called, and that causes a crash.. If you use notifications to add the text that shouldn't happen. I have posted some code that shows how to post notifications between threads in `next-prog' with the subject `Reliably adding NSTimers to an NSRunLoop in another thread.' good luck erik From Joe.Loda at usmail.mpct.com Wed May 26 11:37:21 1999 From: Joe.Loda at usmail.mpct.com (Loda, Joe) Date: Thu Nov 3 14:42:05 2005 Subject: YBOS-2.0 Usage Message-ID: Hi all, Has anyone had any experience using YBOS-2.0 on the shipping version of WebObjects 4.0 (or 4.0.1)? The doc seems to indicate that it was written for DR2. Thanks, Joe --- Joe Loda, mpct Solutions Corporation, Chicago joe.loda@mpct.com (ASCII, MIME) From sanguish at digifix.com Wed May 26 15:14:12 1999 From: sanguish at digifix.com (Scott Anguish) Date: Thu Nov 3 14:42:05 2005 Subject: Is this a multi-threaded problem? Message-ID: <199905262210.SAA00248@digifix.com> > > I am taking the precaution of using an NSLock around the actual > > insertion of text into the textview (the insertReBoldText: for > > example goes through this... although it sets up a bunch of stuff > > first..) > > I don't think a lock is enough because you might still interupt > layout managers that work in the background. The only really safe > solution is to go through an NSConnection. > I had hoped that this wasn't going to be necessary due to the 'threadsafeness' of the AppKit in OSXS... Is this still true? > > > The current problem that I'm finding is that if a user cancels the > > action mid stream, that it can take a substantial time to return > > because the thread is still executing.. if the document gets closed > > before the thread finishes (cancelled or not) the stuff in the > > detached thread that puts the information into the textview still > > gets called, and that causes a crash.. > > If you use notifications to add the text that shouldn't happen. I > have posted some code that shows how to post notifications between > threads in `next-prog' with the subject `Reliably adding NSTimers to > an NSRunLoop in another thread.' > Thanks.. I'll have a look! -- Scott Anguish Middlebury College - Center for Educational Technology Stepwise Server - Mac OS X Server Information From sanguish at digifix.com Thu May 27 00:08:12 1999 From: sanguish at digifix.com (Scott Anguish) Date: Thu Nov 3 14:42:05 2005 Subject: Is this a multi-threaded problem? Message-ID: <199905270704.DAA02473@digifix.com> > > I am taking the precaution of using an NSLock around the actual > > insertion of text into the textview (the insertReBoldText: for > > example goes through this... although it sets up a bunch of stuff > > first..) > > I don't think a lock is enough because you might still interupt > layout managers that work in the background. The only really safe > solution is to go through an NSConnection. > I'm still having crashes here... What I've done at this point is this each document creates an NSConnection that it can serve.. this is done during the -init serverConnection=[[NSConnection defaultConnection] retain]; [serverConnection enableMultipleThreads]; [serverConnection setRootObject:self]; serverConnectionName = [[[NSString allocWithZone:[self zone]] initWithFormat:[[NSProcessInfo processInfo] globallyUniqueString]] retain]; [serverConnection registerName:serverConnectionName]; likewise, when the document deallocs, the connection is invalidated [serverConnection invalidate]; [serverConnection release]; In my method that is detached as a new thread, I do this theConnection= [[NSConnection connectionWithRegisteredName:serverConnectionName host:nil] retain]; selfInMain=[[theConnection rootProxy] retain]; [selfInMain setProtocolForProxy:@protocol(InsertProgressTextProtocol)]; and then whenever I pass text back to the main thread, this is done.. [selfInMain insertRedBoldText:@"\n---- Cancelled by user ----\n"]; which is part of the protocol InsertProgressTextProtocol However, the problem still exists that if the document is closed before the task in the other thread is completed... this should invalidate the serverConnection in the main thread, however theConnection in the other thread appears to be still valid... if ([theConnection isValid]) { [selfInMain insertRedBoldText:@"\n---- Cancelled by user ----\n"]; } this still attempts to execute, even when the serverConnection is invalidated (it then goes ahead and dies, trying to insert the str into a freed object.. All this NSTask handling in the second thread is in a series of loops, so there isn't a way to really interrupt it easily (for notification that that the thread is dead). I've tried checking to make sure that they connection is valid before calling anything that is in selfInMain, but that doesn't seem to be making any difference... Aside from perhaps handling my NSTask in a more complicated manner (reading from the pipes via notifications in the seperate thread instead of just stalling through the loop... ) So I guess I'm still needing help.. > If you use notifications to add the text that shouldn't happen. I > have posted some code that shows how to post notifications between > threads in `next-prog' with the subject `Reliably adding NSTimers to > an NSRunLoop in another thread.' Yep.. I grabbed that.. At this point I've added alot of complexity, but have gained no additional stability than I had earlier today.. I'm somewhat convinced now that the InfoPanel crasher I was dealing with is actually related to the NSDocument textInput stuff that was a problem the other day... Adding that fix, before making any of these other changes, fixed that crasher.. So I consider this to be a two steps forward, one step back situation.. :-) -- Scott Anguish Middlebury College - Center for Educational Technology Stepwise Server - Mac OS X Server Information From Marcus.Miss at commerzbank.com Thu May 27 01:35:54 1999 From: Marcus.Miss at commerzbank.com (Miss, Marcus) Date: Thu Nov 3 14:42:05 2005 Subject: What's wrong with my modal-dialog ? Message-ID: <493AAA8758AAD21182A000A0C9925DE404C40E@sv004212.exchange.commerzbank.com> I have a problem which drives me crazy. My modal dialog under Win/NT is loosing its modal-state after activating another app and reactivating my own app. What I'm doing wrong ? Marcus From F.Knobloch at inis.de Thu May 27 02:56:42 1999 From: F.Knobloch at inis.de (Frank Knobloch) Date: Thu Nov 3 14:42:05 2005 Subject: What's wrong with my modal-dialog ? Message-ID: >I have a problem which drives me crazy. >My modal dialog under Win/NT is loosing its modal-state after activating >another app and reactivating my own app. >What I'm doing wrong ? We have had a similar problem and it occurs also if the screen-saver is active, but there is a patch avaiable: Look at: http://www.hksys.com/~jeremy/openstep/index.html Sincerely Frank --- //InIS GmbH * Loesungen fuer Automations- und Verwaltungssysteme [Addr.: Friedrich-Ebert-Str. 78 ~ D-34119 Kassel] [Tel..: ++49-561-788090, Fax: ++49-561-78 80 929] [Mail.: F.Knobloch@inis.de, Web: www.inis.de] From Philippe.Robert at uptime.ch Thu May 27 04:17:16 1999 From: Philippe.Robert at uptime.ch (Philippe Robert) Date: Thu Nov 3 14:42:05 2005 Subject: Video Performance Message-ID: <9905271117.AA10205@california.uptime.ch> Hi! While porting NXBench to Mac OS X Server 1.1 I noticed, that my G3/350 is twice as slow as my PII/266 with a Millenium II AND the Omni Fast Video Driver 1.0 I use at work. What's the reason behind this? I thought, the Rage128 would be blazingly fast(er)??? sweet dreams, Phil PS: I will upload NXBench for Mac OS X Server in the next view days. -- Philippe C.D. Robert | Uptime ObjectFactory Inc Unix/OpenStep Software Engineer | http://www.uptime.ch http://www.nice.ch/~phip | info@uptime.ch From Marcus.Miss at commerzbank.com Thu May 27 07:37:39 1999 From: Marcus.Miss at commerzbank.com (Miss, Marcus) Date: Thu Nov 3 14:42:05 2005 Subject: TCL + YB Message-ID: <493AAA8758AAD21182A000A0C9925DE404C411@sv004212.exchange.commerzbank.com> I want to use TCL in my application. Works fine except the fact that i'm unable to redirect the stdout. I tried several ways but I failed. Who can help me ? Thanks, Marcus From dix.lorenz at joice.net Thu May 27 08:05:39 1999 From: dix.lorenz at joice.net (Dix Lorenz) Date: Thu Nov 3 14:42:05 2005 Subject: getting legacy code over to MacOSX? Message-ID: Hi, I have on a cvs-Server a hierarchy of source files, about 4 folders deep. I got MacOSX to connect to the server and check out the sources, so now they are nicely on my machine. But when I want to use them in Project Builder, it copies them. Is there any way to prevent that? Or do I have to make a link in my project directory to every single header and source file (> 400 files)? Thanks, Dix From pierce at twinforces.com Thu May 27 09:50:51 1999 From: pierce at twinforces.com (Pierce T. Wetter III) Date: Thu Nov 3 14:42:05 2005 Subject: getting legacy code over to MacOSX? In-Reply-To: References: Message-ID: At 8:18 AM -0700 5/27/99, Dix Lorenz wrote: >Hi, > >I have on a cvs-Server a hierarchy of source files, about 4 folders deep. I >got MacOSX to connect to the server and check out the sources, so now they >are nicely on my machine. But when I want to use them in Project Builder, it >copies them. Is there any way to prevent that? Or do I have to make a link >in my project directory to every single header and source file (> 400 >files)? No, isn't it annoying, and Yes! ProjectBuilder currently insists on doing this (though perhaps if you edited the make file it wouldn't?) Apple said at WWDC they were going to stop this in a future revision of PB, which would be a good thing. However, you can make "sub projects" in each folder, and it won't (though you'll have to have a project file in each folder instead.) There's also a X11/OpenBSD command: "lndir" that creates symbolic links for a whole directory, but you'll have to port it to MOSXS as I haven't gotten around to it yet. You can get the source from: ftp://www.darwinlinux.com/pub/openbsd/usr.bin/lndir Oh, and there's another choice: Don't use ProjectBuilder, use bsdmake or make (make is GNUmake on MOSXS) and vi/emacs instead. That's what I'm doing to port the Open/Net/FreeBSD sources, rather then making 4,000 PB projects. Pierce ---------------------------------------------------------------- Pierce T. Wetter III, Director, Twin Forces, Inc. e-mail: pierce@twinforces.com Phone:520-779-4227 U.S. Mail: 1300 South Milton Rd, Suite 206, Flagstaff, AZ 86001 From paulrs at lgs-systems.com Thu May 27 10:03:28 1999 From: paulrs at lgs-systems.com (Paul R. Summermatter) Date: Thu Nov 3 14:42:05 2005 Subject: NSColorWells in an NSTableView In-Reply-To: <199905251035.KAA00257@listserv.ny.verio.net> References: <199905251035.KAA00257@listserv.ny.verio.net> Message-ID: <9905271703.AA01545@slab> You wrote: [ > You wrote: [ > [ Anybody have any good suggestions for getting an NSColorWell [ into an [ > [ NSTableView? Most other widgets work great, since thery're based [ [ > around cells, but NSColorWell doesn't have a corresponding [ [ > NSColorWellCell. [ > [ > John et al, [ > [ > This is annoying. However, there may have been a good [ reason for [ > it. At first, I thought maybe the cell was just hidden, but, at [ least on OS [ > 4.2 Mach, I can find no special color cell class in the AppKit [ class dump. [ > In any event, it looks to me like some work has been done on the [ > NSColorPanel, so maybe one of the engineers will open up or fix [ this API for [ > the OS X release. [ [ Remember that color wells rely on drag & drop and cells cannot [ register themselves as dragging destination. This makes a cell based [ color well rather unattractive. If you don't need drop support you [ can listen for NSColorPanelColorDidChangeNotifications and draw a [ little coloured area yourself. Erik, What? Cells don't inherit from views and, as such, don't receive drawing calls directly, yet they are expected to draw, and the controls that manage them are responsible for telling them when and where to draw. Don't you think the shelf in the Workspace Manager is a control with a bunch of cells? We have built a number of controls which allow drag and drop and have multiple cells. Much like the drawing api, our controls simply forward the drag & drop messages to the appropriate cell. Regards, Paul --- Paul Summermatter LGS Systems, Inc. Medical Computing Division 15 TJ Gamester Ave Portsmouth, NH 03801-5871 (603) 433-9822 voice (603) 433-9818 fax (888) 898-6321 pager 8986321@skytel.com paging email paulrs@lgs-systems.com (NeXT or MIME Mail Welcome) http://www.lgs-systems.com From nhoj at cd.chalmers.se Thu May 27 09:08:43 1999 From: nhoj at cd.chalmers.se (nhoj@cd.chalmers.se) Date: Thu Nov 3 14:42:05 2005 Subject: Video Performance In-Reply-To: <9905271117.AA10205@california.uptime.ch> References: <9905271117.AA10205@california.uptime.ch> Message-ID: <19990527160844.20190.qmail@glenn> Phil wrote: > Hi! > > While porting NXBench to Mac OS X Server 1.1 I noticed, that my G3/350 is > twice as slow as my PII/266 with a Millenium II AND the Omni Fast Video > Driver 1.0 I use at work. > > What's the reason behind this? I thought, the Rage128 would be blazingly > fast(er)??? > > sweet dreams, Phil > > PS: I will upload NXBench for Mac OS X Server in the next view days. Graphics performance under OPENSTEP/Mac OS X is likely to be limited more by CPU performance and the bandwidth of CPU to PCI device transfers than by the performance of the particular card. >From the performance figures, I would guess that MacOS X Server does not do properly optimized PCI bus transactions. [The Omni Fast Video driver typically improves speed by 2-3 times, and does this by combining framebuffer writes using a feature of the P6 core, allowing efficient long bursts over the PCI bus.] As to the speed of the card, 2D performance hasn't really improved all that much since the Matrox Millenium was released, IIRC. Development seems to be focused on 3D performance instead. Regards, John Hornkvist From maury at OAAI.COM Thu May 27 13:53:53 1999 From: maury at OAAI.COM (Maury Markowitz) Date: Thu Nov 3 14:42:05 2005 Subject: Permissions under NT Message-ID: <199905272049.QAA04270@OAAI.COM> Whenever I get stuck like this it's typically because I'm missing something in the frameworks, so here goes... We want the permissions of our documents to be reasonable, so on the OS-X side we use umask to get the current user mask and then use NSFileManager to set them. However there's no umask under NT (obviously), how do I do something reasonable there? Is there a better solution that works on both? Oh yeah, and why isn't there a "read only" version of umask? We just want to know what the permission are, not change them. Maury From dyoung at vviuh221.vvi.com Thu May 27 14:14:13 1999 From: dyoung at vviuh221.vvi.com (David Young) Date: Thu Nov 3 14:42:05 2005 Subject: Permissions under NT References: <199905272049.QAA04270@OAAI.COM> Message-ID: <9905272114.AA02480@vviuh221.vvi.com> > Oh yeah, and why isn't there a "read only" version of umask? We > just want to know what the permission are, not change them. Is that for mac os x server? Maybe you want to use access() ? Thanks A Bunch! David Young; VVI-DCS dyoung@vvi.com ACCESS(2) UNIX Programmer's Manual ACCESS(2) NAME access - determine accessibility of file SYNOPSIS #include /* POSIX applications #include */ #define R_OK 4/* test for read permission */ #define W_OK 2/* test for write permission */ #define X_OK 1/* test for execute (search) permission */ #define F_OK 0/* test for presence of file */ int access(const char *path, int mode); DESCRIPTION The access function checks the accessibility of the file whose name is given by path for the file access permissions indicated by mode. The value of mode is either the bitwise inclusive OR of the access permissions (R_OK, W_OK and X_OK), or the existence test (F_OK), which cause access to check whether the direc- tories leading to the file can be searched and the file exists. The real user ID and the group access list (including the real group ID) are used in verifying permission, so this call is useful to set-UID programs. Notice that only access bits are checked. A directory may be indicated as writable by access but an attempt to open it for writing will fail (although files may be created there); a file may look executable, but execve(2) will fail unless it is in proper format. RETURN VALUE If path cannot be found or if any of the desired access modes would not be granted, then a -1 value is returned and errno is set to indicate the error. Otherwise, a value of zero is returned. ERRORS If any of the following conditions occurs, access returns -1 and sets errno to the corresponding value: [EACCES] Permission bits of the file mode do not per- mit the requested access, or search permis- sion is denied on a component of the path prefix. The owner of a file has permission checked with respect to the ``owner'' read, write, and execute mode bits, members of the file's group other than the owner have per- mission checked with respect to the ``group'' mode bits, and all others have permissions checked with respect to the ``other'' mode bits. [EFAULT] The path argument points outside the process's allocated address space. [EINVAL] The pathname contains a character with the high-order bit set. [EIO] An I/O error occurred while reading from or writing to the file system. [ELOOP] Too many symbolic links were encountered in translating the pathname. [ENAMETOOLONG] A component of path exceeds 255 characters, or the entire pathname exceeds 1023 charac- ters. For POSIX applications these values are given by the constants {NAME_MAX} and {PATH_MAX}, respectively. [ENOENT] The named file does not exist or path points to an empty string. [ENOTDIR] A component of the path prefix is not a directory. [EROFS] Write access is requested for a file on a read-only file system. [ETXTBSY] Write access is requested for a pure pro- cedure (shared text) file that is being exe- cuted. SEE ALSO chmod(2), stat(2) or stat(2P) From dyoung at vviuh221.vvi.com Thu May 27 14:18:02 1999 From: dyoung at vviuh221.vvi.com (David Young) Date: Thu Nov 3 14:42:05 2005 Subject: Video Performance References: <19990527160844.20190.qmail@glenn> Message-ID: <9905272118.AA02494@vviuh221.vvi.com> nhoj@cd.chalmers.se wrote: > I would guess that MacOS X Server does not do > properly optimized PCI bus transactions. I doubt that. With identical computers and 64Mbytes of RAM Windows NT, Mac OS X Server (Rhapsody/Intel), and OpenStep/Mach have basically the same general purpose drawing rates as substantiated by our real-time draw validation programs. With Mac OS X Server/PPC the general purpose draw rate is basically the same taking into account the clock frequency of the CPU. I suggested to sweet dreams Phil that a 2X factor can be related to a video color depth setting 12 v.s. 24 bit depth. Without knowing exact details it is hard to tell. Also, just for a complete comparison: general purpose drawing rate on Solaris is pathetically slow. Thanks A Bunch! David Young; VVI-DCS dyoung@vvi.com From alexnet at gestalt.com Thu May 27 14:45:23 1999 From: alexnet at gestalt.com (Alex Molochnikov) Date: Thu Nov 3 14:42:05 2005 Subject: Connecting G3 to Ethernet Thin cable Message-ID: <9905272145.AA00659@gestalt.com> A non-text attachment was scrubbed... Name: not available Type: text/enriched Size: 316 bytes Desc: not available Url : /mailman/archive/macosx-dev/attachments/19990527/a77fb38b/attachment.bin From michaelmonner at home.com Thu May 27 19:38:41 1999 From: michaelmonner at home.com (Michael Monner) Date: Thu Nov 3 14:42:05 2005 Subject: Connecting G3 to Ethernet Thin cable Message-ID: <19990528023841.OEOO14995.mail.rdc1.on.home.com@ToneServer> >Does anyone know a way of connecting a G3 box to the Ethernet Thin cable? We have coax >connections in our network, and Apple does not provide a transceiver that would connect it to >10BaseT socket in G3. I was told you couldn't go from 10baseT to 10base2 when I got my imac so I changed my network. With the G3 I suppose you could get an ethernet card with a thin port and ignore the built-in port. Michael From alexnet at gestalt.com Thu May 27 19:57:02 1999 From: alexnet at gestalt.com (Alex Molochnikov) Date: Thu Nov 3 14:42:05 2005 Subject: Connecting G3 to Ethernet Thin cable Message-ID: <9905280257.AA00951@gestalt.com> Michael Monner wrote: > With the G3 I suppose you could get an ethernet card with a thin port and ignore the built-in port. I wish it were that simple. Third-party network cards require drivers, and for MacOSXS there aren't any. I've spent half a day today on the phone talking to different manufacturers, only to hear their lamentantions about Apple not giving them the necessary specs. Apple's tech support were not of great help either, mostly confirming that I am up the S. creek. However, I found a sort of a gender-bender that takes 10BaseT connector at one end, and coax cable connector (BNC) at the other. Unfortunately, both are female outlets, so I will have to complement it with a couple of male-male ending cables. I wonder if there is something that I should be aware of before trying to use these (electric impedance? just guessing...) In the meantime, thanks to all those who responded to my cry in the wilderness. Alex Molochnikov Phoenix Data Trend -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: text/enriched Size: 993 bytes Desc: not available Url : /mailman/archive/macosx-dev/attachments/19990527/29bff050/attachment.bin From bbum at codefab.com Thu May 27 20:16:24 1999 From: bbum at codefab.com (Bill Bumgarner) Date: Thu Nov 3 14:42:05 2005 Subject: Connecting G3 to Ethernet Thin cable In-Reply-To: <9905280257.AA00951@gestalt.com> Message-ID: Why not pick up an el-cheapo four port hub w/a coax connector as the uplink port?? I picked one up that does exactly this for less than $80-- and that was over four years ago!! If not that, you can always get a hub w/one of the DB-XX/AUI ports on it, and pick up an AUI<->Coax adaptor for $25 or so. b.bum On Thu, 27 May 1999, Alex Molochnikov wrote: > Michael Monner wrote: > > > With the G3 I suppose you could get an ethernet card with a thin port and > ignore the built-in port. > > I wish it were that simple. Third-party network cards require drivers, and > for MacOSXS there aren't any. I've spent half a day today on the phone > talking to different manufacturers, only to hear their lamentantions about > Apple not giving them the necessary specs. Apple's tech support were not of > great help either, mostly confirming that I am up the S. creek. > > However, I found a sort of a gender-bender that takes 10BaseT connector at > one end, and coax cable connector (BNC) at the other. Unfortunately, both are > female outlets, so I will have to complement it with a couple of male-male > ending cables. I wonder if there is something that I should be aware of > before trying to use these (electric impedance? just guessing...) > > In the meantime, thanks to all those who responded to my cry in the wilderness. > > Alex Molochnikov > Phoenix Data Trend > From wchin at acm.org Thu May 27 20:46:52 1999 From: wchin at acm.org (Bill Chin) Date: Thu Nov 3 14:42:05 2005 Subject: Connecting G3 to Ethernet Thin cable Message-ID: <199905280251.WAA08360@mail.richmond.net> > > With the G3 I suppose you could get an ethernet card with a thin port and ignore > > the built-in port. > > I wish it were that simple. Third-party network cards require drivers, and for > MacOSXS there aren't any. I've spent half a day today on the phone talking to > different manufacturers, only to hear their lamentantions about Apple not > giving them the necessary specs. Apple's tech support were not of great help > either, mostly confirming that I am up the S. creek. Actually, the DEC21x4Ethernet driver in MOXS works with quite a few cards. For instance, I had a Kingston KNE100TX (no drivers for Mac, no OpenFirmware) in my Blue and Ice G3 running w/o a problem for a while (running NetBoot too). Of course, MacOS didn't work with it... I've seen reports of cross-platform Asante and Znyx cards working... basically, if you can verify that the card has a DEC 21140 chipset, it _should_ work. To be sure, I'd get it from someplace you can return it to. I had a Cogent EM590 lying around somewhere with a DEC 21040 chipset... I can't find it to check compatibility though. ..Bill Chin From ahoesch at smartsoft.de Fri May 28 01:43:59 1999 From: ahoesch at smartsoft.de (Andreas =?iso-8859-1?Q?H=F6schler?=) Date: Thu Nov 3 14:42:05 2005 Subject: Connecting G3 to Ethernet Thin cable In-Reply-To: References: Message-ID: <9905280844.AA01868@smartsoft.de> A non-text attachment was scrubbed... Name: not available Type: text/enriched Size: 171 bytes Desc: not available Url : /mailman/archive/macosx-dev/attachments/19990528/717f1ac5/attachment.bin From Philippe.Robert at uptime.ch Fri May 28 01:32:15 1999 From: Philippe.Robert at uptime.ch (Philippe C.D. Robert) Date: Thu Nov 3 14:42:05 2005 Subject: Video Performance In-Reply-To: <19990527160844.20190.qmail@glenn> References: <19990527160844.20190.qmail@glenn> Message-ID: <9905280832.AA00349@uptime.ch> You wrote: > Graphics performance under OPENSTEP/Mac OS X is likely to be limited more by > CPU performance and the bandwidth of CPU to PCI device transfers than by the > performance of the particular card. Sorry, it was all my fault - I didn't saw that the screen depths weren't equal when I made the tests.... sweet dreams, Phil --- Philippe C.D. Robert Uptime Object Factory Inc http://www.nice.ch/~phip Software Development http://www.projectcenter.ch OpenStep/Unix From sbrandon at music.gla.ac.uk Fri May 28 03:08:12 1999 From: sbrandon at music.gla.ac.uk (Stephen Brandon - SysAdmin) Date: Thu Nov 3 14:42:05 2005 Subject: Bugfix for [window cacheImageInRect:] Message-ID: <199905281008.LAA09189@music.gla.ac.uk> A non-text attachment was scrubbed... Name: not available Type: text/enriched Size: 1682 bytes Desc: not available Url : /mailman/archive/macosx-dev/attachments/19990528/56547c46/attachment.bin From nhoj at cd.chalmers.se Fri May 28 02:16:39 1999 From: nhoj at cd.chalmers.se (nhoj@cd.chalmers.se) Date: Thu Nov 3 14:42:05 2005 Subject: Video Performance In-Reply-To: <9905272118.AA02494@vviuh221.vvi.com> References: <9905272118.AA02494@vviuh221.vvi.com> Message-ID: <19990528091640.23589.qmail@glenn> David Young wrote: > nhoj@cd.chalmers.se wrote: > > > I would guess that MacOS X Server does not do > > properly optimized PCI bus transactions. > > I doubt that. With identical computers and 64Mbytes of RAM > Windows NT, Mac OS X Server (Rhapsody/Intel), and OpenStep/Mach > have basically the same general purpose drawing rates as substantiated by > our real-time draw validation programs. With Mac OS X Server/PPC the > general purpose draw rate is basically the same taking into account the > clock frequency of the CPU. Sounds good. Has anyone tried DeadLift on MacOS X Server? Just for clarification, where these machines P6-core (PPro, PII, Celeron PIII) systems, and was the OmniFastVideo driver running on the OpenStep/Mach system? Also, exactly what do your real-time draw validation programs do? I.e. are they rendering limited, or bandwidth limited? Regards, John Hornkvist From mike at lorax.com Fri May 28 09:46:45 1999 From: mike at lorax.com (Mike Ferris) Date: Thu Nov 3 14:42:05 2005 Subject: NSDocument/NSTextView crasher Message-ID: <199905281742.KAA03187@boom.lorax.com> Let me add that this fix is appropriate for Scott's app where his NSDocument subclass is the window's delegate. If you use a NSWindowController subclass as well, and it is the window delegate, then the method below should go there. Actually, since the method is a notification observer, it can go anywhere as long as the object it's in observes the NSWindowWillCloseNotification. The main point is that if the NSTextView is not the first responder of the window when it closes, this bug will be avoided. Mike Begin forwarded message: > From: Scott Anguish > Date: 1999-05-26 01:22:16 -0700 > To: Multiple recipients of list > Subject: Re: NSDocument/NSTextView crasher > X-Nextstep-Mailer: Mail 4.2mach [i386] (Enhance 2.2p2) > X-Listprocessor-Version: 6.0 -- ListProcessor by Anastasios Kotsikonas > Originator: macosx-dev@omnigroup.com > X-Comment: To unsubscribe, follow directions at > http://www.omnigroup.com/MailArchive/ > > I noticed that this wasn't posted by Mike to the list.. so I figured I'd > forward it (since it is important).. > > Andrew's fix caused a leak... this doesn't.. > > Thanks again to Mike and Andrew (and everyone else who sent comments) > > Mike said: > > This is a known problem that has been fixed in OSX DP1 (it doesn't happen to > you there, does it?) > > A safe workaround that will not cause leaks would be to add the following to > your Document class in the test case: > > - (void)windowWillClose:(NSNotification *)notification { > [[notification object] makeFirstResponder:[notification object]]; > } > > The problem is that the NSInputManager is not noticing that the NSTextView > goes away, so getting the text view to give up first responder makes it notice. > This will work before and after the fix. > From raoul at mobopro.com Fri May 28 12:10:10 1999 From: raoul at mobopro.com (Raoul Duke) Date: Thu Nov 3 14:42:05 2005 Subject: SCSI drivers? Message-ID: <374EEA12.3338BAC9@mobopro.com> Apple's support and information websites are quite confusing to me. After some weird searching ("scsi and mac and os and x and server") I learned that the only SCSI cards that were tested by Apple for use with Mac OS Server (and my G3/BW/300) are the: Adaptec 2940 (Single-Channel Ultra2 LVD SCSI) Adaptec 2930 (Ultra SCSI) I'd like to purchase the 2930 from Adaptec, but can't find the driver for this. Furthermore, I can't even find Apple's OEM'd version from the Apple Store. Any ideas? --Raoul From raoul at mobopro.com Fri May 28 13:03:12 1999 From: raoul at mobopro.com (Raoul Duke) Date: Thu Nov 3 14:42:05 2005 Subject: SCSI drivers? References: <374EEA12.3338BAC9@mobopro.com> Message-ID: <374EF680.9024E71A@mobopro.com> Raoul Duke wrote: > > Apple's support and information websites are quite confusing to me. > After some weird searching ("scsi and mac and os and x and server") I > learned that the only SCSI cards that were tested by Apple for use with > Mac OS Server (and my G3/BW/300) are the: > > Adaptec 2940 (Single-Channel Ultra2 LVD SCSI) > Adaptec 2930 (Ultra SCSI) > > I'd like to purchase the 2930 from Adaptec, but can't find the driver > for this. Furthermore, I can't even find Apple's OEM'd version from the > Apple Store. Any ideas? An update. I phoned Apple Store and they informed me they don't sell the SCSI card separately, and I should have added on the $49 option at the time. I was fooled by the marketing of the clever easy-access door, I guess, and assumed you could install components later without bother! :-) They recommended I phone Apple Technical Support and see what they recommended. After I directed them to TIL article 24888, Support suggested two options: a) buying the Adaptec 2930U and simply trying it, or b) calling an Apple authorized service provider and seeing if they'd sell/install me one as a repair. Granted option b) would be substantially more expensive than the original BTO option, but it might be the only way. Support also seemed to think a driver for OS X Server is not the issue, but getting the proper firmware was. Not sure what to do next.... --Raoul From raoul at mobopro.com Fri May 28 13:52:27 1999 From: raoul at mobopro.com (Raoul Duke) Date: Thu Nov 3 14:42:05 2005 Subject: SCSI drivers? References: <199905281949.PAA04037@mail.richmond.net> Message-ID: <374F020B.BA2F1A07@mobopro.com> Bill Chin wrote: > > This really isn't for macosx-dev. Search the macosx-admin archives > and you'll see that this topic has been hashed over. Oops, didn't think of that. Sorry! --Raoul From raoul at mobopro.com Sat May 29 14:29:06 1999 From: raoul at mobopro.com (Raoul Duke) Date: Thu Nov 3 14:42:05 2005 Subject: Workspace Manager content inspector Message-ID: <37505C22.CD8C1D5A@mobopro.com> Is it possible to install a custom Workspace Manager content inspector, as under NEXTSTEP 3.3 via subclassing WMInspector? Thanks, Raoul From richard at brainstorm.co.uk Sun May 30 01:53:54 1999 From: richard at brainstorm.co.uk (Richard Frith-Macdonald) Date: Thu Nov 3 14:42:05 2005 Subject: drag-and-drop example? Message-ID: <9905300853.AA00495@tiptree.brainstorm.co.uk> This is probably a faq somewhere, but I don't know where to look. Can anyone point me to simple example source for drag-and-drop? I'm thinking of a pair of trivial apps, one with a view acting as a drag source and the other with one acting as a drag destination for some sort of data. --- `Everything should be made as simple as possible, but not simpler.' From wjs at omnigroup.com Sun May 30 19:54:01 1999 From: wjs at omnigroup.com (William Shipley) Date: Thu Nov 3 14:42:05 2005 Subject: Another warning from a co-moderator. Message-ID: <199905310254.TAA08530@ignem.omnigroup.com> While there's been a lot of good content recently, I feel compelled to make an announcement to certain miscreants: This mailing list is for asking and answering questions about developing software for Mac OS X. It's such a simple concept. Why, why, why can't everyone follow it? Insulting people's user interfaces: Nope. Right out. Even a child, I think, can see that this doesn't qualify as a question about developing software. Speculating about how cool it would be if someone would port a new filesystem to Darwin. Sing it along with me, you folks in back: "That's not a question about developing software!" Speculating about how such a filesystem port might be done: No! No, no, no! It's just noise, folks. We can spend all day talking about what might be the right way to write software. Here's the test: if you aren't writing the software, and you're not answering someone's question about the software, you are contributing to the noise. I know you can counter and say, "Well, if someone WERE to write X piece of code, they'll be greatly helped by this nugget of widsom." That may be true, but the point is, if this hypothetical person ever does write the hypothetical code, and hypothetically asks a question, THEN it would be appropriate for you to share your knowledge. Until then, even if you think your opinion is clever and helpful and will generally benifit the programming community at large, it simply DOESN'T BELONG HERE. Announcing special pricing on drawing tools: C'mon, Andy, just because you're my friend doesn't mean you can post that here. -- I'm not interested in any rules-lawyer interpretations of this list's charter, either. I know that, given enough levels of abstraction, ANYTHING qualifies as being of interest to developers. ("I have a new recipe for lasagne, and I KNOW developers have to eat, or they'll die, so I thought I'd post it...") I am heartened to see so many people actually getting use out of this list. I am heartened actual Apple engineers are still reading and addressing our concerns. Those of you contributing to the noise, be warned that you are risking banishment. Yours most cordially, -Wil Shipley Founder of the "I Liked Jar-Jar Binks Club" From tom at evatac.com Sun May 30 21:00:36 1999 From: tom at evatac.com (Tom Carstensen) Date: Thu Nov 3 14:42:05 2005 Subject: external code modules Message-ID: <1283999233-51910272@evatac.com> How should one load and call an external code module for Carbon/OS X? Normally, for 68K i directly call the code via a proc ptr, and under PPC i create a UPP. Since NewRoutineDescriptor and CallUniveralProc are not defined in the Carbon API (when i link these are missing), i tried call it directly as i would with 68K code: MyProcPtr procPtr; code = GetResource('CODE', 128); HLock(code); procPtr = (MyProcPtr) *code; (*procPtr)(.....); HUnlock(code); This crashes under carbon, works fine in 68K. Any ideas? Thanks -Tom Carstensen From ahoesch at smartsoft.de Mon May 31 01:52:47 1999 From: ahoesch at smartsoft.de (Andreas Hoschler) Date: Thu Nov 3 14:42:05 2005 Subject: DO problem on MacOS X Server Message-ID: <9905310852.AA00832@smartsoft.de> Hi, I am just porting a client server system from OPENSTEP to MacOS X Server. The clients use DO (NSConnection,...) to communicate with the server. Running the server on an OPENSTEP machine and the client on MacOS X Server works, vice versa does not (The registered connection is not found). Running both (client and server) on the same MacOS X Server machine does work too (I unfortunately have only one Mac yet,so...). Any ideas? Did I miss anything regarding DO on MacOS X Server? Thanks a lot, Andreas From paulrs at lgs-systems.com Mon May 31 07:10:30 1999 From: paulrs at lgs-systems.com (Paul R. Summermatter) Date: Thu Nov 3 14:42:05 2005 Subject: DO problem on MacOS X Server In-Reply-To: <9905310852.AA00832@smartsoft.de> References: <9905310852.AA00832@smartsoft.de> Message-ID: <9905311410.AA00606@slab> You wrote: [ Hi, [ [ I am just porting a client server system from OPENSTEP to MacOS X [ Server. The clients use DO (NSConnection,...) to communicate with the [ server. Running the server on an OPENSTEP machine and the client on [ MacOS X Server works, vice versa does not (The registered connection is [ not found). Running both (client and server) on the same MacOS X Server [ machine does work too (I unfortunately have only one Mac yet,so...). [ [ Any ideas? Did I miss anything regarding DO on MacOS X Server? [ [ Thanks a lot, [ [ Andreas [ Andreas et al, This is interesting. I just did some tests like this where the server was running on YB/WOF 4.0 NT and the OpenStep clients could connect. Haven't tried it yet on MOSX, but I will try to do so this week and let you know how I make out. Regards, Paul --- Paul Summermatter LGS Systems, Inc. Medical Computing Division 15 TJ Gamester Ave Portsmouth, NH 03801-5871 (603) 433-9822 voice (603) 433-9818 fax (888) 898-6321 pager 8986321@skytel.com paging email paulrs@lgs-systems.com (NeXT or MIME Mail Welcome) http://www.lgs-systems.com From cb at df.lth.se Mon May 31 07:43:07 1999 From: cb at df.lth.se (Christian Brunschen) Date: Thu Nov 3 14:42:05 2005 Subject: DO problem on MacOS X Server In-Reply-To: <9905311410.AA00606@slab> Message-ID: On Mon, 31 May 1999, Paul R. Summermatter wrote: > You wrote: > [ Hi, > [ > [ I am just porting a client server system from OPENSTEP to MacOS X > [ Server. The clients use DO (NSConnection,...) to communicate with the > [ server. Running the server on an OPENSTEP machine and the client on > [ MacOS X Server works, vice versa does not (The registered connection is > [ not found). Running both (client and server) on the same MacOS X Server > [ machine does work too (I unfortunately have only one Mac yet,so...). > [ > [ Any ideas? Did I miss anything regarding DO on MacOS X Server? Did you enable Remote Mach IPC on the Mac OS X Server machine? The same way you have to arrange things to accept remote NSHost connections; DO also runs over Mach IPC, I beleive ? > [ > [ Thanks a lot, > [ > [ Andreas > [ > > Andreas et al, > > This is interesting. I just did some tests like this where the > server was running on YB/WOF 4.0 NT and the OpenStep clients could connect. > Haven't tried it yet on MOSX, but I will try to do so this week and let you > know how I make out. > > Regards, > Paul // Christian Brunschen From rcfa at cubiculum.com Mon May 31 09:33:03 1999 From: rcfa at cubiculum.com (Ronald C.F. Antony) Date: Thu Nov 3 14:42:05 2005 Subject: Sorry for bouncing e-mail Message-ID: <199905311633.MAA00985@kannix.cubiculum.com> For about one hour there was an e-mail outage here, during which messages sent to me were returned with "user unknown" errors. [*] Mail service is again fully restored. So please, don't unsubscribe me from the list or something like that. If anyone sent me a personal message, please resend it. Thanks, Ronald [*][This was the accidental by product of debugging lookupd (for more info on that subject see the macosx-admin@omnigroup.com list).] ============================================================================== "The reasonable man adapts himself to the world; the unreasonable one persists in trying to adapt the world to himself. Therefore all progress depends on the unreasonable man." G.B. Shaw | rcfa@cubiculum.com | NeXT-mail welcome From mike at lorax.com Mon May 31 10:15:21 1999 From: mike at lorax.com (Mike Ferris) Date: Thu Nov 3 14:42:05 2005 Subject: DO problem on MacOS X Server Message-ID: <199905311810.LAA07308@boom.lorax.com> > Any ideas? Did I miss anything regarding DO on MacOS X Server? Yes, something changed. There's a release note on this somewhere, but basically cross-host Mach port lookup is disabled by default in OSXServer for security reasons. You can turn it back on (the release note says how...) This is the reason it works when client and server both run on the same box (ie that's not cross-host), but when the server runs on an OSXS box, and you have not enabled this, remote clients cannot connect. Mike Ferris Begin forwarded message: > From: Andreas Hoschler > Date: 1999-05-31 01:04:26 -0700 > To: Multiple recipients of list > Subject: DO problem on MacOS X Server > X-Nextstep-Mailer: Mail 4.2mach (Enhance 2.1) > X-Listprocessor-Version: 6.0 -- ListProcessor by Anastasios Kotsikonas > Originator: macosx-dev@omnigroup.com > X-Comment: To unsubscribe, follow directions at > http://www.omnigroup.com/MailArchive/ > > Hi, > > I am just porting a client server system from OPENSTEP to MacOS X Server. > The clients use DO (NSConnection,...) to communicate with the server. Running > the server on an OPENSTEP machine and the client on MacOS X Server works, > vice versa does not (The registered connection is not found). Running both > (client and server) on the same MacOS X Server machine does work too (I > unfortunately have only one Mac yet,so...). > > Any ideas? Did I miss anything regarding DO on MacOS X Server? > > Thanks a lot, > > Andreas > From maury at OAAI.COM Mon May 31 12:29:03 1999 From: maury at OAAI.COM (Maury Markowitz) Date: Thu Nov 3 14:42:05 2005 Subject: A couple more NT porting questions Message-ID: <199905311923.PAA12665@OAAI.COM> Some UI issues... a) is there a standard command key for closing a window? Some use Control-W, is that correct? Is there also another, like Alt-F3? b) where exactly should the Close item go? IB has it in the Window menu, some apps have it in File just above Exit, others have it in File near the top of the menu. c) how do I indicate "backspace" in the menu item for "Clear" (in edit)? d) is Control-Q really used on Windows? Is there some way to specify Alt-F4? e) how do I make a letter the "key" for that menu item, for instance I want to make the T in Tools the key for that menu - is it &T? Another issue... I'm having problems with the Services menu. When you hit it, even passing by (on the validate I'm guessing) an error comes up on the Console about _fastCString (sorry, I'm on the OS-X box here) and the app stops responding. I used a new copy of the menu from the menu palette in IB, but the same thing happens. I'm not doing anything special to the menu, do I have to? Maury From ari at dsimulation.com Mon May 31 06:44:34 1999 From: ari at dsimulation.com (Ari Danesh) Date: Thu Nov 3 14:42:05 2005 Subject: Running Standard Java Message-ID: <00034c27b043cbfd_mailit@relay.macconnect.com> I am trying to invoke the command line "java" to run and application that has been compiled and jarred in Windows NT. I have put this application on my Mac (Blue Box) and other NT machines with no problems, but when I try to run it on my MacOS Server I get a can't find class error. Do I have to set up the paths somewhere? Ari From pf at ipagents.com Mon May 31 14:17:25 1999 From: pf at ipagents.com (Peter Fagerlund) Date: Thu Nov 3 14:42:05 2005 Subject: Running Standard Java In-Reply-To: <00034c27b043cbfd_mailit@relay.macconnect.com> Message-ID: Create a Java directory in /Local/Library/ so You now got a directory /Local/Library/Java/ here U place Your jars and classes and the JVM will find them. I also sometimes creates a run.sh text file containing : -- #!/bin/sh export CLASSPATH CLASSPATH=classes:$JAVASCOPE_HOME/JavaScope.zip:infobus.jar java sun.beanbox.BeanBoxFrame -- and in the my dev directory - terminal - issue a # sh run.sh after the JavaScope.zip: You place your added classes and jars /peter At 22.50 +0200 99-05-31, Ari Danesh wrote: >I am trying to invoke the command line "java" to run and application that has >been compiled and jarred in Windows NT. I have put this application on my Mac >(Blue Box) and other NT machines with no problems, but when I try to run it >on my MacOS Server I get a can't find class error. Do I have to set up the >paths somewhere? > > >Ari From sanguish at digifix.com Mon May 31 15:17:10 1999 From: sanguish at digifix.com (Scott Anguish) Date: Thu Nov 3 14:42:05 2005 Subject: Stupid SplitView problem Message-ID: <199905312217.SAA21809@digifix.com> I can't believe this is something I have to ask.. but here it goes.. I've got a splitview in a window, and I'm storing the position that it is saved as.. (no problem so far).. How do I restore it when I open the window again? I've tried just setting the frame on the contents of one side of the splitview, but that doesn't seem to work.. -- Scott Anguish Middlebury College - Center for Educational Technology Stepwise Server - Mac OS X Server Information From fedor at ultra.doc.com Mon May 31 20:52:37 1999 From: fedor at ultra.doc.com (Adam Fedor) Date: Thu Nov 3 14:42:05 2005 Subject: RFV: GNUstep development tasks Message-ID: <37535904.4C2384D8@ultra.doc.com> GNUstep needs volunteers to finish off a few tasks in preparation for the upcoming 0.6.0 release. If you have the necessary skills and have from 10-30 hours of time in the next few months, please consider doing some of these tasks. It would be very appreciated and help get GNUstep a lot further towards it's goal of becoming a world-class developement environment in the spirit of OpenStep and MacOSX. 1. Integrate changes to the NSText class into the current libraries (Note: the changes have already been written, they just need to be integrated in and tested). 2. Talk to Window Maker developers about what needs to be done to better integrate with GNUstep. 3. Finish off drag and drop support 4. Add support for installing into a shallow install tree. 5. Write a simple installer app (or modify an existing one). 6. Certify compatibility with OpenStep and MacOSX class documentation. 7. Fix up a few classes that are somewhat lacking... Please see http://www.gnustep.org/developers/tasks.html for more information. We would especially appreciate help from experienced developers, but even less experienced poeple can work on some of the tasks. -- Adam Fedor | Those who can't do, simulate Digital Optics Co. | fedor@doc.com (MIME) |------------------------------ fedor@gnu.org (GNUstep!) From F.Schippers at inter.NL.net Mon May 31 22:25:00 1999 From: F.Schippers at inter.NL.net (Frans H. Schippers) Date: Thu Nov 3 14:42:05 2005 Subject: Creator and File types Message-ID: Hi i want to be able to get and set the Creator and File-type of Mac-files under MacOS X Server. Has anyone some pointers for documentation. With A/UX you had a program setfile i believe? Frans Frans H. Schippers mailto:F.Schippers@inter.nl.net Grote Beer 189 tel: +31 20 44 54 117 1188 AZ Amstelveen mob: +31 65 36 50 806 Netherlands From sanguish at digifix.com Mon May 31 23:17:32 1999 From: sanguish at digifix.com (Scott Anguish) Date: Thu Nov 3 14:42:05 2005 Subject: Still more splitview woes Message-ID: <199906010618.CAA23436@digifix.com> I got the general message that you need to implement the delegate method - (void)splitView:(NSSplitView *)sender resizeSubviewsWithOldSize:(NSSize)oldSize And I'm doing that now.. however I"m still having just totally bizzare results.. in the "size" inspector in InterfaceBuilder, what should the spring/autosize settings be for the NSBoxes that are automatically and used as the super view for the left/right or top/bottom areas of the slip views? The default seems to be no springs, so Its probably safe to assume that this is managed elsewhere.. But here, with no autosize settings.. and using this code... when it gets called with the value not equal to -1 (just use the settings in the nib) its using the standard adjustSubviews, and that seems to be OK... but Well... I"m just swinging widly at this point... anyone got a working example to munch on? - (void)splitView:(NSSplitView *)sender resizeSubviewsWithOldSize:(NSSize)oldSize { if (sender == annotationsSplitView) { NSLog(@"---------------------------------------\nannotationsSplitView resizeSubviewsWithOldSize:"); if ([lessonPage pageAnnotationSplitViewHeight] == -1.0) [sender adjustSubviews]; else { // we need to make the pageAnnotationSplitView the correct height; NSRect splitViewFrame; NSRect lowerViewFrame; NSRect upperViewFrame; splitViewFrame=[annotationsSplitView frame]; upperViewFrame=[upperAnnotationSplitViewBox frame]; lowerViewFrame=[lowerAnnotationSplitViewBox frame]; upperViewFrame.size.height=[lessonPage pageAnnotationSplitViewHeight]; lowerViewFrame.size.height=NSHeight(splitViewFrame)-upperViewFrame.size.height-[sender dividerThickness]; [upperAnnotationSplitViewBox setFrame:upperViewFrame]; [lowerAnnotationSplitViewBox setFrame:lowerViewFrame]; } } } thanks -- Scott Anguish