From gerard at ina.fr Mon Feb 1 01:42:02 1999 From: gerard at ina.fr (Gerard Iglesias) Date: Thu Nov 3 14:41:48 2005 Subject: problem tracking mouse drag in NSCell References: Message-ID: <36B576EA.6A09C972@ina.fr> Anders Pytte wrote: > > I override NSCell trackMouse, loop, call nextEvent requesting mouse moved > events, draw feedback, and call setNeedsDisplayInRect on the control view > after each draw. Overriding trackMouse mean that you are making your own event loop, hence the display induced by setNeedsDisplayInRect can't be achieved. > But this does not work consistently unless I also call > controlView display after calling setNeedsDisplayInRect, in which case my > code works fine. It is absolutely normal. If you want to use setNeedsDisplayInRect you can override the methods: startTrackingAt:inView: continueTracking:at:inView: stopTracking:at:inView:mouseIsUp:, in place of setNeedsDisplayInRect. But in this case you will not be able to manage mouse moved event. About this subject, I think that system, in the YB, that permits to manage mouse moved event is too complex. It was necessary when the computer was not powerful, but now it seems to be obsolete a bit. Sincerely. Gerard From lavoie at cst.ca Mon Feb 1 06:25:48 1999 From: lavoie at cst.ca (Martin-Gilles) Date: Thu Nov 3 14:41:48 2005 Subject: DR2 on new G3s Message-ID: <199902011424.JAA15073@plexus.cst.ca> >Is the Prelude to Rhapsody CD essentially Openstep 4.2? I have been >having difficulty getting upgrading from Openstep 4.1 to Openstep 4.2 on >my intel machine. If it is indeed 4.2 or so I would love to find it >somehow also. Prelude is an actual OpenStep 4.2 FAT CD. By downloading the boot disk image for black hardware (it only came with Intel boot disk), I was actually able to upgrade my used color slab from NeXTSTEP 3.3 to OpenStep 4.2. Though, I do miss RenderMan. Martin-Gilles Lavoie | "No! Try not. Mac OS programmer, CS&T | Do, or do not. lavoie@cst.ca | There is no try." http://blackhole.cst.ca/ | -- Yoda on error handling From maury at OAAI.COM Mon Feb 1 08:05:46 1999 From: maury at OAAI.COM (Maury Markowitz) Date: Thu Nov 3 14:41:48 2005 Subject: development idea Message-ID: <199902011200.HAA07514@OAAI.COM> > Look at MailViewer.app (in the WebObjects menu): it works as either > a pop client or with standard Unix mail. Hardly a replacement in this context though. With the exception of a few nicities when dealing with MIME there's not much going for it. I've spoken on this topic before. CyberDog is a _very_ good place to start when looking at mail clients. Although it had two really big drawbacks - it was very slow and stored mail in a propietary format - the interface and functionality was rich. For instance the same editor window was used for both reading and writing mail and news, both sources were treated as "mailboxes", and you could make filters that could be applied to either. Everything was indexed (but no server-side newsgroup searches, GRRR) and searchable, it handled MIME OK (not great though) etc. That aside the proprietary format of CD's mail store is an issue. A xxx to MailViewer (or more specifically, mbox format) is likely the first thing that people will need. Maury From jean-michel.cazaux at finindev.com Mon Feb 1 09:04:03 1999 From: jean-michel.cazaux at finindev.com (Jean-Michel Cazaux) Date: Thu Nov 3 14:41:48 2005 Subject: Functions & Methods with variable number of arguments ??? Message-ID: <01BE4E0D.414A43C0@JEAN-MICHEL> Hi folks, I'm looking for tips or samples of code to write functions ( like NSLog ) and methods (like [NSArray arrayWithObjects:...]) with variable number of arguments. I also want the functions to be able to have NSObjects arguments, do I have to put the source in a C file ? 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 guenther at onevision.de Mon Feb 1 10:42:00 1999 From: guenther at onevision.de (Guenther Fuerthaller) Date: Thu Nov 3 14:41:48 2005 Subject: Functions & Methods with variable number of arguments ??? References: <01BE4E0D.414A43C0@JEAN-MICHEL> Message-ID: <199902011842.TAA17058@onevision.de> A non-text attachment was scrubbed... Name: not available Type: text/enriched Size: 506 bytes Desc: not available Url : /mailman/archive/macosx-dev/attachments/19990201/8459a718/attachment.bin From janos.lobb at yale.edu Mon Feb 1 10:56:48 1999 From: janos.lobb at yale.edu (=?iso-8859-1?Q?J=E1nos?= =?iso-8859-1?Q?_?= =?iso-8859-1?Q?L=F6bb?=) Date: Thu Nov 3 14:41:48 2005 Subject: DR2 & ATTO SCSI Message-ID: Hi, I made the mistake trying to install DR2 on a partition of an internal HD connected to SCSI bus 2 on a wide ATTO card. /Only the CD is on the built-in internal SCSI of the PowerMac 9500/120/. Booting from the CD was OK but the install reported that there is no hard drive to install to. /On the wide SCSI there is 22GByte waiting to be commanded by DR2/. Reading the manual... I recognized that DR2 is unable to deal with additional SCSI cards. After exiting from the installer I rebooted the machine but that 1Gig partition did not show up. Disk First Aid and Drive Setup did not see it either. The HDs were formated by Hard Disk Toolkit, so I ran it, but HD toolkit did not show it either. Norton 4.0 also did not show it in the "missing drives" list. Zapped the PROM and rebuilt the desktop. After it HD toolkit showed the partition with a question mark but was unable to mount it. The partition is shown as three or four separate Apple format partitions. Question is how can I regain that 1Gig as a normal Mac HFS volume. J?nos P.S. I tried to put BEOS 3.0 into the same partition hoping it will bring it back, but wonder... BEOS also cannot deal with the Wide ATTO SCSI just like DR2. J?nos L?bb Tel: 203-737-5204 Yale University Pathology Fax: 203-785-7303 310 Cedar St. Room BML104A janos.lobb@yale.edu New Haven CT 06510 Never take a candid-cookie from a stranger. From Scott_Harrison at apple.com Mon Feb 1 12:33:46 1999 From: Scott_Harrison at apple.com (Scott Harrison) Date: Thu Nov 3 14:41:48 2005 Subject: Help wanted: "Timing" problem with insertObject: in EOEditingContext In-Reply-To: <19990128222852.C375@bright.nu> Message-ID: <9902011533.AA463242@mage> Hi all, I do not know if anyone has answered this yet, as I might have misplaced some mail... Daniel Brahneborg made my mailer see: -> Hi again, -> -> It turned out to be a typo on my part that caused the problem with the -> derived classes, that I wrote about a few days ago. Sorry about that. :) -> -> Today I found another strange problem, that I can't quite figure out, -> related to the database interface. -> -> In my application I sometimes create new objects to be inserted into the -> database, and often those objects have to create a few other objects to -> stay consistent. In a manual to WebObjects, they do something like this: -> -> Class* object = [[[Class alloc] init] autorelease]; -> AnotherClass* object2 = [[[AnotherClass alloc] init] autorelease]; -> [object addObject:object2 toBothSidesOfRelationshipWithKey:@"relation"]; -> [editContext insertObject:object]; -> [editContext insertObject:object2]; -> -> So... What is the "correct" way to create new database objects? Is the -> order I've found correct, or is there any other better way to do it? -> One should create the objects, insert them into the editing context, and then create the relationships. The act of inserting the objects into the editing context helps to ensure that the to-many relationships will work because the arrays will be correct. Therefore: Class *object = [[[Class alloc] init] autorelease]; AnotherClass *object2 = [[[AnotherClass alloc] init] autorelease]; [editingContext insertObject: object]; [editingContext insertObject: object2]; [object addObject: object2 toBothSidesOfRelationshipWithKey: @"relation"]; The question of whether one should use autorelease versus using release at a different time is sometimes one of style. When one can say: Class *object = [[Class alloc] init]; // other stuff here that still only uses object locally [object release]; then usually it is better to not use autorelease since it is generally more efficient not to put an object into an autorelease pool. In creating a Company, Department and Address, the relationships between them should not influence the generation of the primary key. If you need to create two address instances, and insert each into the editing context (with code like that above), the normal EOF mechanisms should generate two unique primary keys (1 for each object). If you are manually creating primary keys for the address instances you might run into trouble. If you have your primary keys as class properties I would suggest you change that as soon as possible (assuming you have the capability to influence the schema). By the way, if you want only one Address for each Company/Department you can always access the Address across the relationship between the two. For example, keep the Address in the Department and access it through department.address. Or if you want to have a company address and a department address which can be the same or different, you can always start them off the same by doing something like: Company *c = [[[Company alloc] init] autorelease]; Department *d = [[[Department alloc] init] autorelease];; Address *a = [[[Address alloc] init] autorelease]; [ec insertObject: c]; [ec insertObject: d]; [ec insertObject: a]; [c setAddress: a]; [d setAddress: a]; [c addObject: d toBothSidesOfRelationshipWithKey: @"department"]; // Note that we are assuming that there is no reverse relationship // from Address to either Department or Company -> Thanks a lot for your help. -> -> /Basic -> Scott Harrison From Armin.Tenge at prakinf.tu-ilmenau.de Mon Feb 1 14:10:29 1999 From: Armin.Tenge at prakinf.tu-ilmenau.de (Armin Tenge) Date: Thu Nov 3 14:41:48 2005 Subject: from ObjectiveC to Java Message-ID: Hi, moving from ObjC to Java is not so easy. I cannot comprehend why Apple didn't deliver the DeveloperTutorial in Java and a tutorial from ObjC to Jav. Can anyone help me? Thanks, Armin From ajetha at tacticalstep.com Mon Feb 1 14:39:57 1999 From: ajetha at tacticalstep.com (Alykhan Jetha) Date: Thu Nov 3 14:41:48 2005 Subject: from ObjectiveC to Java References: Message-ID: <36B62D3C.FC98E767@tacticalstep.com> I can't help on the tutorial, but for understanding Java from C/ObjC point of view, I would recommend a copy of O'Reilly's JAVA in Nutshell. While it compares Java to C and C++ and not ObjC, it helped me tremendously. ISBN: 1-56592-262-X Armin Tenge wrote: > Hi, > moving from ObjC to Java is not so easy. I cannot comprehend why Apple > didn't deliver the DeveloperTutorial in Java and a tutorial from ObjC > to Jav. Can anyone help me? > > Thanks, > Armin -- Alykhan Jetha (AJ) TacticalStep Inc., Email: mailto:ajetha@tacticalstep.com Phone: 905-831-2643 ext 1 Fax: 905-837-2546 From eb at object.com Mon Feb 1 14:45:27 1999 From: eb at object.com (Eric Bergerson) Date: Thu Nov 3 14:41:48 2005 Subject: Functions & Methods with variable number of arguments ??? In-Reply-To: <01BE4E0D.414A43C0@JEAN-MICHEL> References: <01BE4E0D.414A43C0@JEAN-MICHEL> Message-ID: <9902012245.AA26192@object.com> Jean-Michel, Below, is the "_main" from a quick tool I wrote to demonstrate the three basic ways of using varargs. BTW, if you are on Mach, you can just say "man varargs" on the command line to see some basic UNIX examples. As far as writing functions vs. methods, they behave exactly the same, as I illustrate below. One way of using varargs, is simply to gather the structure the varargs are managed in, and pass them to another method or function that is designed to use that structure. This is demonstrated in the method below myOwnNSLog:. If you do not want to simply use another method expecting variable arguments, but rather, would like to process the variable arguments yourself, you need to decide how you will determine how many arguments there are. There is no way I know of to get the number of arguments provided. So, you need to choose one of a couple of different approaches. One approach is to explicitely provide the number of variable arguments. The method below, logObjectsOfCount:description: uses the "count" argument to determine how many arguments there are. Clearly if the argument count incorrectly describes the number of arguments, this method could fail. Another way is to use some kind of identifier to delimit the end of the list of variable arguments. In the function logObjectsOfCount(), the function determines when the arguments are done by checking for the first nil argument. Obviously, this will fail if the caller does not provide a nil at the end of the list of arguments. Also, this approach is no good if it is possible for one of the arguments to be validly nil, which would terminate the list early. Of course, in the case of a function like NSLog(), the number of variable arguments are determined by parsing the "format" provided, to determine what arguments and types are expected. Interestingly, I never discovered how to get the compiler to give my vararg methods similar warnings to the wonderful ones that methods like stringWithFormat: get, if the variable arguments don't match the format. If you need any further assistance, please don't hesitate to call or write. Sincerely, Eric Bergerson Objective Technologies, Inc. email: eb@object.com voice: 212 988 6268 fax: 212 988 6273 web: http://www.object.com LGS Systems, Inc. email: eb@lgs-systems.com web: http://www.lgs-systems.com #import @interface TestObj: NSObject + (void) logObjectsOfCount: (int) count description: (NSString *) format, ...; + (void) myOwnNSLog: (NSString *) format, ...; @end @implementation TestObj + (void) logObjectsOfCount: (int) count description: (NSString *) format, ... { va_list ap; id nextObject = nil; int i; va_start(ap,format); for(i=0; i%@",num1, num2, str3, nil); [TestObj logObjectsOfCount: 3 description:@"method-->%@",num1, num2, str3]; [TestObj myOwnNSLog: @"Num1: %@ Num2: %@ Str3: %@", num1, num2, str3]; [pool release]; exit(0); // insure the process exit status is 0 } -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: text/enriched Size: 4120 bytes Desc: not available Url : /mailman/archive/macosx-dev/attachments/19990201/5aace9dc/attachment.bin From karl at ensuing.com Mon Feb 1 18:54:32 1999 From: karl at ensuing.com (Karl Kraft) Date: Thu Nov 3 14:41:48 2005 Subject: Subclassing Inspectors in IB (on Windows) Message-ID: <000f01be4e57$5c650c40$3c03040a@kiva.ensuing.net> On OS/NS I've developed a subclass of NSButton called ActiveButton. This button style changes images as you pass over it 'ala Explorer, my browser of choice. :-) Since I want this to be a strong reusable object, I've developed a palette and an inspector for the object in InterfaceBuilder, and subclased IBButtonInspector, the inspector for NSButtons. This allows me to inherit all the functionality of the regular NSButton Inspector. (header courtesy of class-dump) The palette compiles great on Rhapsody and OpenStep, but fails on Windows during the link. The problem is the referenece to IBButtonInspector, which is compiled into InterfaceBuilder, and not available in any framework. ActiveButtonInspector.o : error LNK2001: unresolved external symbol .objc_class_name_IBButtonInspector One possible solution would to change the superclass of ActiveButtonInspector to something else, and then clone the class to a subclass of IBButtonInspector at runtime, but this seems the "difficult" way to do it. Does anyone have and "easy" solution to this problem? From cwhite at comnetix.com Mon Feb 1 20:25:40 1999 From: cwhite at comnetix.com (Craig White) Date: Thu Nov 3 14:41:48 2005 Subject: Subclassing Inspectors in IB (on Windows) In-Reply-To: <000f01be4e57$5c650c40$3c03040a@kiva.ensuing.net> Message-ID: <9902012325.AA402001@deli> >ActiveButtonInspector.o : error LNK2001: unresolved external symbol >.objc_class_name_IBButtonInspector We use the following in Makefile.preamble to 'relax' the linker... OTHER_LDFLAGS = -undefined suppress From karl at ensuing.com Mon Feb 1 20:44:42 1999 From: karl at ensuing.com (Karl Kraft) Date: Thu Nov 3 14:41:48 2005 Subject: Subclassing Inspectors in IB (on Windows) Message-ID: <000801be4e66$bff09b80$3c03040a@kiva.ensuing.net> Indeed, this solved the problem I was having, which brings up question #2: Where can one find documentation on the MS linker? -----Original Message----- From: Craig White To: Multiple recipients of list Date: Monday, February 01, 1999 8:32 PM Subject: Re: Subclassing Inspectors in IB (on Windows) > >>ActiveButtonInspector.o : error LNK2001: unresolved external symbol >>.objc_class_name_IBButtonInspector > >We use the following in Makefile.preamble to 'relax' the linker... > >OTHER_LDFLAGS = -undefined suppress > > From gerard at ina.fr Tue Feb 2 00:24:59 1999 From: gerard at ina.fr (Gerard Iglesias) Date: Thu Nov 3 14:41:49 2005 Subject: Subclassing Inspectors in IB (on Windows) References: <9902012325.AA402001@deli> Message-ID: <36B6B65B.51EDDFEE@ina.fr> Craig White wrote: > > >ActiveButtonInspector.o : error LNK2001: unresolved external symbol > >.objc_class_name_IBButtonInspector > > We use the following in Makefile.preamble to 'relax' the linker... > > OTHER_LDFLAGS = -undefined suppress In fact there is an error in the bundle makefile : Apple\Developer\Makefiles\pb_makefiles\bundle.make ifeq "$(OS)" "NEXTSTEP" PROJTYPE_LDFLAGS = $($(OS)_PROJTYPE_LDFLAGS) -bundle -undefined suppress endif ifeq "$(OS)" "WINDOWS" PROJTYPE_LDFLAGS = $($(OS)_PROJTYPE_LDFLAGS) -bundle endif You can add definitely the "-undefined suppress" options at the end of the fifth line. Sincerely. Gerard From gerard at ina.fr Tue Feb 2 00:27:38 1999 From: gerard at ina.fr (Gerard Iglesias) Date: Thu Nov 3 14:41:49 2005 Subject: Subclassing Inspectors in IB (on Windows) References: <000801be4e66$bff09b80$3c03040a@kiva.ensuing.net> Message-ID: <36B6B6FA.4170A9A7@ina.fr> Karl Kraft wrote: > > Indeed, this solved the problem I was having, which brings up question #2: > > Where can one find documentation on the MS linker? > I think that it is not a problem of the MS linker, it is the same for the linker on the Mach platform, see the bundle.make file. Sincerely. Gerard From gerard at ina.fr Tue Feb 2 00:33:39 1999 From: gerard at ina.fr (Gerard Iglesias) Date: Thu Nov 3 14:41:49 2005 Subject: from ObjectiveC to Java References: Message-ID: <36B6B863.B1BCC5A@ina.fr> Armin Tenge wrote: > > Hi, > moving from ObjC to Java is not so easy. I cannot comprehend why Apple > didn't deliver the DeveloperTutorial in Java and a tutorial from ObjC > to Jav. Can anyone help me? > Well, a good solution would to stay with ObjC :-) I know, I know... it is not very contructive, but I couldn't resist :-) Sincerely. Gerard. PS: I hate the "pensee unique" around Java. "pensee unique" : a French expression for speaking about the thing that caracterize that everybody have to think like the others people: liberalism, Java, C++, Windows... From stark at easynet.fr Tue Feb 2 08:11:11 1999 From: stark at easynet.fr (stark@easynet.fr) Date: Thu Nov 3 14:41:49 2005 Subject: Functions & Methods with variable number of arguments ??? References: <9902012245.AA26192@object.com> Message-ID: <9902021611.AA06719@easynet.fr> Hi, > + (void) myOwnNSLog: (NSString *) format, ... > { > va_list ap; > NSString *aDescription; > > > va_start(ap,format); > aDescription = [[[NSString alloc] initWithFormat: format > arguments: ap] autorelease]; > va_end(ap); > > NSLog(aDescription); > } You should always write _two_ methods when dealing with varargs: * The one that get the '...' (ie: the argument are in its stack frame). It sould call: * The other one that take a 'va_list' argument (Ie: it takes a pointer to the stack frame of another method) This way you can reuse your varadic functions. (I strongly recommend to do this. The day you need it, it is very handy to have it without having to dig again in the sources modifying working things, or doing copy/paste) Example: This is what clients uses: - (void)almLogError:(NSString *)aFormat,... { va_list argp; va_start( argp, aFormat ); [self almLogError:aFormat arguments:argp]; va_end( argp ); } This is what does the job: - (void)almLogError:(NSString *)aFormat arguments:(va_list)argList { NSString *theMessage = [[NSString alloc] initWithFormat:aFormat arguments:argList]; [itsLogFile writeData:[theMessage dataUsingEncoding:[theMessage smallestEncoding]]]; [itsLogFile writeData:[@"\n" dataUsingEncoding:[@"\n" smallestEncoding]]]; [theMessage release]; } We can reuse the almLogError:arguments: for other things: - (void)almLogWarning:(NSString *)aFormat,... { if ([self almIsVerbose]) { va_list argp; va_start( argp, aFormat ); [self almLogError:aFormat arguments:argp]; va_end( argp ); } } Or: - (void)almFatal:(NSString *)aFormat,... { va_list argp; va_start( argp, aFormat ); [self almLogError:aFormat arguments:argp]; va_end( argp ); [NSException raise:@"ALMFatalError" format:@"See log file for details"]; } (If you do pure "C" functions, the convention is to add a 'v' before the name of the function. For instance: printf/vprintf, sprintf/vsprintf, fprintf/vfprintf, NSLog/NSLogv oops. Think different. ) Cheers, --fred From ash at sineware.com Tue Feb 2 10:14:18 1999 From: ash at sineware.com (Ashish Mishra) Date: Thu Nov 3 14:41:49 2005 Subject: from ObjectiveC to Java Message-ID: <199902021814.NAA07894@engmail.uwaterloo.ca> > from ObjectiveC to Java > moving from ObjC to Java is not so easy. I cannot comprehend why Apple > didn't deliver the DeveloperTutorial in Java and a tutorial from ObjC > to Jav. Can anyone help me? I did manage to do all the tutorials using the Java bridge, I don't know if that's what you want to do. The biggest problem was the lack of documentation on the exposed Yellowbox APIs. But that documentation is now on Apple's dev site. If you have any specific questions, I can probably give you a hand. -- Ashish Mishra, E&CE Dept, University of Waterloo From randy at oceansoft.com Tue Feb 2 12:48:15 1999 From: randy at oceansoft.com (Randy Leonard) Date: Thu Nov 3 14:41:49 2005 Subject: Common code-base for OpenStep & YellowBox Message-ID: <9902022048.AA15816@oceansoft.com> A non-text attachment was scrubbed... Name: not available Type: text/enriched Size: 517 bytes Desc: not available Url : /mailman/archive/macosx-dev/attachments/19990202/48caa9b3/attachment.bin From paulrs at lgs-systems.com Tue Feb 2 13:59:11 1999 From: paulrs at lgs-systems.com (Paul R. Summermatter) Date: Thu Nov 3 14:41:49 2005 Subject: EOTextAssociation Message-ID: <9902022159.AA00648@slab> Re: EOF 2.2, OS 4.2 Mach & NT Hey guys, I'm hoping someone out there has already been through this problem or has a good understanding of the text system on OS 4.2 and can give me some insight on how to work around this problem. For some reason, I don't know if it's the fault of EOTextAssociation or NSTextView or what, there is a rather nasty way to lose data that's been entered into an NSTextView in an EOApplication. If you are editing information in a textView and you switch to another window and make changes to an object causing (possibly by saving those changes) the EOTextAssociation in your original window to get a subjectChagned message, rather than recognizing that the user is in the middle of entering data, everything that the user just typed is WIPED out. I watched the association, and it seems to recognize that the text has changed (ie. it sends the displayGroup an associationDidBeginEditing message). However, unlike the EOGenericControlAssociation, rather than just leaving the text in the middle of editing, it appears that it decides to 'refresh' the text view from the eo's value. Should I subclass EOTextAssociation and handle this on my own? Is this fixed in EOF 3.0? Is there anything different about the NSTextView object (from an NSControl) which would make it impossible for the EOTextAssociation to leave the edit alone? Any thoughts/tips/hints would be well appreciated. 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 leigh at cs.uwa.edu.au Tue Feb 2 20:59:17 1999 From: leigh at cs.uwa.edu.au (Leigh Smith) Date: Thu Nov 3 14:41:49 2005 Subject: Speculation: More objective C like Java? In-Reply-To: <199902021725.JAA04592@ignem.omnigroup.com> References: <199902021725.JAA04592@ignem.omnigroup.com> Message-ID: <199902030459.MAA01639@antechinus.cs.uwa.edu.au> This is merely my speculation, but I thought I'd test the waters. I love Objective C. Actually, what I really love is the object syntax, the ability to pass methods (using selectors), and categories. Admittedly left-of-field, but if a "new, improved" Java that had ObjC named parameters (perhaps without the parameter ordering requirement, as in CLOS), compiled to native code and introduced a category/selector mechanism existed, are there any other features missing that would be needed for the bulk of ObjC programmers to see this "Objective-Java" in a more favourable light and make coding in it more compelling? Obviously there are the semantics of C pointers, but this could be replaced with some memory access object. There is also the dynamicism of ObjC that is still to be achieved in Java. Obviously there is Sun's protectionism of the name that may make the point moot. Perhaps this could be achieved with a preprocessor, without requiring much rework of the existing Java compilers, I guess I should look into David Stes' POC again... The way I see it is Apple could do to Java what NeXT did to C, retain compatibility with a standard, but raise the paradigm up a notch. Not meaning to start a war, but hoping to find holes in my (modest?) proposal. Email replies may be the best approach. Regards --- Leigh Computer Music Lab, Computer Science Dept, Smith University of Western Australia +61-8-9380-2279 leigh@cs.uwa.edu.au (NeXTMail/MIME) Microsoft - What do you want to re-install today? From rcfa at cubiculum.com Tue Feb 2 21:29:24 1999 From: rcfa at cubiculum.com (Ronald C.F. Antony) Date: Thu Nov 3 14:41:49 2005 Subject: Speculation: More objective C like Java? In-Reply-To: <199902030459.MAA01639@antechinus.cs.uwa.edu.au> References: <199902030459.MAA01639@antechinus.cs.uwa.edu.au> Message-ID: <9902030529.AA10039@kannix.cubiculum.com> > are there any other features missing that would be > needed for the bulk of ObjC programmers to see this "Objective-Java" in a > more favourable light and make coding in it more compelling? message forwarding dynamic typing true class methods are also missing, not just syntax, posing and categories.... 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 stark at easynet.fr Wed Feb 3 01:32:02 1999 From: stark at easynet.fr (stark@easynet.fr) Date: Thu Nov 3 14:41:49 2005 Subject: Common code-base for OpenStep & YellowBox References: <9902022048.AA15816@oceansoft.com> Message-ID: <9902030932.AA00347@easynet.fr> Hi, > To all: > > I am having some difficulty compiling a common code base on both YellowBox > and OpenStep. To date, I have converted the makefiles as per Apple's > instructions. I have successfully compiled the common code base on both > OpenStep/NT and YellowBox/NT. Unfortunately, I have had no luck doing the > same for OpenStep/Mach. > > Has anyone had success compiling source code for both YellowBox/NT and > OpenStep/Mach? Any suggestions? Do not use the Apple provided convertion tools. Use "PBUnity" from Eric Sunshine which does the thing correctly (albeit the license is a bit intrusive) Cheers, --fred From paulrs at lgs-systems.com Wed Feb 3 07:05:21 1999 From: paulrs at lgs-systems.com (Paul R. Summermatter) Date: Thu Nov 3 14:41:49 2005 Subject: EOTextAssociation followup Message-ID: <9902031505.AA01171@slab> RE EOF 2.2 OS 4.2 Mach & NT Hey guys, Further tests of the problem on which I posted yesterday are leading me to believe that the fault lies with EOTextAssociation. When an EOTextAssociation gets a subjectChanged message, it seems to ignore the fact that its text view is in the middle of an edit. I set break points on EOTextAssociation on the methods, subjectChanged, endEditing, and updateTextObjectWithValue:, and I found that in subjectChanged, updateTextObjectWithValue: is called without first calling endEditing. Here's the stack: #0 0xf00ae40 in -[EOTextAssociation updateTextObjectWithValue:] () #1 0xf00b17a in -[EOTextAssociation subjectChanged] () #2 0x1e016928 in -[EODelayedObserverQueue notifyObserversUpToPriority:] () #3 0x1e016760 in -[EODelayedObserverQueue _processEndOfEventObservers:] () #4 0x18008bd7 in -[NSObject performSelector:withObject:] () #5 0x1801d648 in -[NSPerformTimer fire] () #6 0x1800a362 in -[NSRunLoop _doPerformersForMode:reclaim:] () #7 0x1801540a in -[NSRunLoop limitDateForMode:] () #8 0x600a9c5 in _DPSNextEvent () #9 0x6016e48 in -[NSApplication nextEventMatchingMask:untilDate:inMode:dequeue:] () #10 0x6021b6e in -[NSApplication run] () #11 0x61c70ae in NSApplicationMain () #12 0x3554 in main (argc=1, argv=0xbffffe38) at MDIQ_main.m:11 #13 0x345e in start () at crt0.c:144 It seemed to me that it ought to be innocuous enough to just call endEditing in subjectChanged. So, I implemented my own subclass (shown below), which seems to have fixed the problem. Unfortunately, because I have added an ivar, I cannot use a poseAsClass to address this problem globally. Regardless, I'm still uncomfortable with this solution, as it seems moronically simple. I'm fearful that there is a good reason why the implementor of EOTextAssociation didn't do exactly what I am doing. If anyone out there could confirm this problem or offer any other suggestions, I would appreciate it. Regards, Paul @implementation LGSTextAssociation - (BOOL)isEditing { return _isEditing; } - (void)subjectChanged { if ([self isEditing]) { [self endEditing]; } return [super subjectChanged]; } - (void)textDidBeginEditing:(NSNotification *)aNotification { _isEditing = YES; } - (void)textDidEndEditing:(NSNotification *)aNotification { _isEditing = NO; } @end From jocke at rat.se Wed Feb 3 10:34:45 1999 From: jocke at rat.se (Joakim Johansson) Date: Thu Nov 3 14:41:49 2005 Subject: : Common code-base for OpenStep & YellowBox (& Rhapsody) [solution] Message-ID: <000801be4fa3$e1b2af40$86d30ec2@baldwin.rat.se> > Has anyone had success compiling source code for both YellowBox/NT and > OpenStep/Mach? Any suggestions? Keep the OpenStep 4.2 makefiles. On DR2, make two links: /NextDeveloper/Makefiles -> /System/Developer/Makefiles /NextLibrary/Frameworks -> /System/Library/Frameworks Then the 4.2 makefiles now just works without conversion. Under Yellow Box CR1 / Windows NT (aka WO4) you can achieve the same functionality by copying the /Apple/Developer directory to /Apple/NextDeveloper. No, it's not nice, it takes a bit of diskspace, but it's very convenient when you want to support 4.2/CR1/DR2/whatever without having to mess around with the makefile conversion scripts continuously. That's all there is to it, works like a charm. I would assume that the OSE/4.2 environment can be patched in a similar manner. On a related note: Have anyone had problems with nib incompatibilites between 4.2 and the WO4 YB runtime? (the application would crash when the nib is loaded. Load it into IB and resave and everything works like perfectly....) Cheers, Joakim -- Joakim Johansson Software Engineer, Research & Trade jocke@rat.se, http://www.rat.se/ From thomas.roehl at cai.com Wed Feb 3 10:42:24 1999 From: thomas.roehl at cai.com (Roehl, Thomas) Date: Thu Nov 3 14:41:49 2005 Subject: Mac OS X Server for developers Message-ID: <94B9DAE20532D2119FD800805F85CABE010C587E@usilmse1.cai.com> I was just curious what people were thinking about Apple's announcement that they are going to charge Select developers for Mac OS X Server. As far as I know, this is the first time that they have charged developers for system software in addition to their developer fees. Is this true? Why would they do this? From kcd at jumpgate.com Wed Feb 3 10:59:59 1999 From: kcd at jumpgate.com (Kenneth C. Dyke) Date: Thu Nov 3 14:41:49 2005 Subject: Mac OS X Server for developers In-Reply-To: <94B9DAE20532D2119FD800805F85CABE010C587E@usilmse1.cai.com> References: <94B9DAE20532D2119FD800805F85CABE010C587E@usilmse1.cai.com> Message-ID: <199902031900.LAA16601@babylon5.jumpgate.com> A non-text attachment was scrubbed... Name: not available Type: text/enriched Size: 578 bytes Desc: not available Url : /mailman/archive/macosx-dev/attachments/19990203/c4fa1888/attachment.bin From btoback at optc.com Wed Feb 3 11:19:14 1999 From: btoback at optc.com (Bruce Toback) Date: Thu Nov 3 14:41:49 2005 Subject: Mac OS X Server for developers Message-ID: <199902031924.MAA12274@landru.optc.com> Thomas Roehl writes: >I was just curious what people were thinking about Apple's >announcement that they are going to charge Select developers >for Mac OS X Server. As far as I know, this is the first time >that they have charged developers for system software in >addition to their developer fees. Is this true? Why would they >do this? MacOS X Server is a completely different OS from MacOS. You won't get charged for MacOS 8.6, which is a very substantial revision from 8.5. I don't know what they'll do for the "full" MacOS X, and I don't mind paying for value received. I just hope that their Enterprise Alliance people won't also have to pay extra. We're in both Select and EAP programs, which is expensive enough. -- 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 epeyton at epicware.com Wed Feb 3 11:18:47 1999 From: epeyton at epicware.com (Eric S. Peyton) Date: Thu Nov 3 14:41:49 2005 Subject: Mac OS X Server for developers In-Reply-To: <94B9DAE20532D2119FD800805F85CABE010C587E@usilmse1.cai.com> Message-ID: All, Don't go jumping any guns - they haven't announce the value of that price yet - so it isn't worth discussion and speculation. This isn't the right place for it, nor is it the right time. Eric epeytoN@epicware.com On Wed, 3 Feb 1999, Roehl, Thomas wrote: > I was just curious what people were thinking about Apple's > announcement that they are going to charge Select developers > for Mac OS X Server. As far as I know, this is the first time > that they have charged developers for system software in > addition to their developer fees. Is this true? Why would they > do this? > > From zmonster at heliosgroup.com Wed Feb 3 13:48:57 1999 From: zmonster at heliosgroup.com (Eric Hermanson) Date: Thu Nov 3 14:41:49 2005 Subject: Compiler/Linker Errors - Unresolved Symbols from external library Message-ID: <5948F7268092D211B2020000D11C0FB8081736@chi-serv2.heliosgroup.com> Hello, I'm using OpenStep 4.2 / NT: I am writing some code to a C API. This API requires that I link in a static library. The documentation (assuming I am a MSVC++ developer) says I need to use the "/MT" compiler option. "/MT" stands for "multi-threaded", and it supposedly does two things: - Defines the _MT symbol - links in LIBCMT.lib (Microsoft's multi-threaded version of libc). So in OpenStep's ProjectBuilder, I manually defined _MT as a compiler option (-D_MT=1), and I added LIBCMT.lib to the Libraries folder (I downloaded this static library from a machine that had MSVC++ installed). The problem is that I'm still getting unresolved symbols from the library that I originally linked in that supports the custom C API I am writing to: librvcst.lib(session.obj) : error LNK2001: unresolved external symbol _strdup librvcst.lib(assoc.obj) : error LNK2001: unresolved external symbol _strdup librvcst.lib(sd_pid.obj) : error LNK2001: unresolved external symbol _getpid _strdup and _getpid are standard functions, but they're not getting resolved due to this multi-threaded situation I describe above. I actually even tried to reference _getpid and _strdup directly in my code, but I still get link errors. Any input would be greatly appreciated. Thank you, Eric From jocke at rat.se Wed Feb 3 14:00:31 1999 From: jocke at rat.se (Joakim Johansson) Date: Thu Nov 3 14:41:49 2005 Subject: Spurious crashes under YB CR1 Message-ID: <000e01be4fc0$9f904560$86d30ec2@baldwin.rat.se> Hi all, I just wanted to check if anyone else have encountered this weird crashing behaviour under YB CR1 (WO4)... We've got a fairly large application that we've ported from NEXTSTEP to OS/YB that are crashing quite inconsistently under YB/NT after a period of use with a quite weird backtrace (we haven't seen these problems under 4.2/Mach or DR2): ----- Program received signal SIGSEGV, Segmentation fault. 0x3100a1ae in objc_msgSend () (gdb) where #0 0x3100a1ae in objc_msgSend () #1 0x529fce in _dpsStat.32 () #2 0x3422499f in NSMenu.o:__i_NSMenu__update () #3 0x342245d4 in -[NSMenu performKeyEquivalent:] () #4 0x3409fce5 in -[NSApplication sendEvent:] () #5 0x3409e941 in -[NSApplication run] () #6 0x340a2074 in NSApplicationMain () #7 0x4eee71 in main (argc=1, argv=0x3320e60) at Orc_main.m:29 #8 0x4efdc5 in WinMain@16 () #9 0x4ef887 in WinMainCRTStartup () ----- The backtrace isn't always exactly the one above, but NSMenu, -performKeyEquivalent: and the inclusion of internal __i_NSMenu.... functions is a common trait... Anyone who has seen anything remotely similar or have any hints on where to start looking? We've been running the application consistently under gdb setting up a breakpoint whenever an exception is raised to rule out that any uncaught exception have screwed things up. I'd say it might be a memory smasher if it weren't for the extremely similar backtraces, who rather points to some weird interaction between our application and the AppKit. Any ideas are welcome! Thanks, Joakim -- Joakim Johansson Software Engineer, Research & Trade jocke@rat.se, http://www.rat.se/ From johnt at healthcc.com Wed Feb 3 16:22:49 1999 From: johnt at healthcc.com (John Turnipseed) Date: Thu Nov 3 14:41:49 2005 Subject: Mac OS X Server for developers In-Reply-To: <94B9DAE20532D2119FD800805F85CABE010C587E@usilmse1.cai.com> Message-ID: At 12:47 PM -0600 02/03/1999, Roehl, Thomas wrote: >I was just curious what people were thinking about Apple's >announcement that they are going to charge Select developers >for Mac OS X Server. As far as I know, this is the first time >that they have charged developers for system software in >addition to their developer fees. Is this true? Why would they >do this? I have no problem with paying for it. What I have a problem with is not being able to test my software under it (Mac OS app - NetBoot & BlueBox) before my clients can buy it and attempt to use it. My clients have the money to afford to just buy something and assume that my software will work with it. -=- John +----------------------------+----------------------------------+ | John Turnipseed | HealthCare Communications, Inc. | | V.P. Product Development | Lincoln, Nebraska | | johnt@healthcc.com | | +----------------------------+----------------------------------+ From sanguish at digifix.com Wed Feb 3 19:43:23 1999 From: sanguish at digifix.com (Scott Anguish) Date: Thu Nov 3 14:41:49 2005 Subject: Mac OS X Server for developers In-Reply-To: <199902031924.MAA12274@landru.optc.com> References: <199902031924.MAA12274@landru.optc.com> Message-ID: <199902040343.WAA16453@digifix.com> > I just hope that their Enterprise Alliance people won't also have to pay > extra. We're in both Select and EAP programs, which is expensive enough. They don't. the FAQ posted yesterday say that Premiere and EAP members get it free of charge. From sanguish at digifix.com Wed Feb 3 21:33:17 1999 From: sanguish at digifix.com (Scott Anguish) Date: Thu Nov 3 14:41:49 2005 Subject: Using NSURL and NSTextView on YB 1.0 (Wof 4.0) Message-ID: <199902040533.AAA16842@digifix.com> I've been playing with NSURL on WOF 4.0 Yellow Box, and found some interesting behaviour... and a question.. if you get make an URL targetURL=[NSURL URLWithString:@"http://www.stepwise.com/"]; and then get the data there... theData=[targetURL resourceDataUsingCache:NO]; and use it to make an NSMutableAttributedString... theParsedHTML=[[[NSMutableAttributedString alloc] initWithHTML:theData baseURL:targetURL documentAttributes:&documentAttributes] autorelease]; and THEN set an NSTextView's text storage to this new string.. [[htmlView textStorage] setAttributedString:theParsedHTML]; You get the HTML to display in the view, as you'd expect. But what you don't expect is that it gradually is loading in all the extra graphics that may be in the document! This is great, but there are two things I need to know... - can I shut this behaviour off? - more importantly, can I get some sort of notification as each of the graphics loads? I've not tried to determine where the loading is taking place as yet.. (I'll be working on that now..) Any help would be greatly appreciated... From ajetha at tacticalstep.com Wed Feb 3 22:44:18 1999 From: ajetha at tacticalstep.com (Alykhan Jetha) Date: Thu Nov 3 14:41:49 2005 Subject: [EOF 3.0] Problems with 'On Demand Locking' Message-ID: <36B941C1.AA7665C5@tacticalstep.com> Folks, [Yellow App] If I set my current editing context to lockObjectBeforeFirstModification:YES, I am unable to trap the exception that is raised when a lock on the object already exists. I've tried both EditingContext's -editingContext:shouldPresentException: and NSApplication's -reportException:, but the suckers escapes me. The OpenBase adaptor seems to work as espected, but the Oracle adaptor attempts to lock objects that are not yet persistent in the database!! (i.e. I haven't saved my changes yet ...) Also, it seems that locks are sensitive to instances of EOEditingContext. (i.e. If I have locked an object in one editingContext and then I pass that object along to another editingContext using -localInstanceOfObject:( in EOUtilities), it generates the appropriate SQL then throws an exception saying that the object is already locked) I can see that this is the desired behavior with WO apps, but it's kinda troublesome in Yellow apps. if ([Apple isListening]) { It would be nice if you could set the locking scope either whole app or editingContext. It would also be nice if the editingContext's delegate got a specific messages in regards to locks (in on-demand mode) - (BOOL)shouldLockObject:(id)anObject; - (void)unableToLockObject:(id)anObject; This way we could handle things in the scope of the particular editingContext. } I tried searching for a delegate method that would be appropriate, but no luck ... If anyone has any hints or tips in regards to this, it would be greatly appreciated! -- Alykhan Jetha (AJ) TacticalStep Inc., Email: mailto:ajetha@tacticalstep.com Phone: 905-831-2643 ext 1 Fax: 905-837-2546 From gclem at frontline-software.dk Thu Feb 4 07:18:33 1999 From: gclem at frontline-software.dk (Geert B. Clemmensen) Date: Thu Nov 3 14:41:49 2005 Subject: NSTableView woes Message-ID: <199902041518.QAA06580@snaps.frontline-software.dk> Hi there, I have an NSTableView to which I programmatically add columns. My problem is that the last column always gets a very small width, no matter what I do with [tableColumn setWidth:]. The total width of the columns is larger than what the clip view "covers", i.e. the horizontal scroller is "active". Anybody with an idea on how to set the width of the last added column. Thx. Geert From markm at mail.tyrell.com Thu Feb 4 07:12:49 1999 From: markm at mail.tyrell.com (Mark F. Murphy) Date: Thu Nov 3 14:41:49 2005 Subject: Mac OS X Server for developers In-Reply-To: <94B9DAE20532D2119FD800805F85CABE010C587E@usilmse1.cai.com> Message-ID: At 10:49 AM -0800 2/3/99, Roehl, Thomas wrote: >I was just curious what people were thinking about Apple's >announcement that they are going to charge Select developers >for Mac OS X Server. As far as I know, this is the first time >that they have charged developers for system software in >addition to their developer fees. Is this true? Why would they >do this? I'm reserving judgement until I see how much they'll be offering it for. Select members don't get QuickTime Pro for free. 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 marco at sente.ch Thu Feb 4 07:17:06 1999 From: marco at sente.ch (Marco Scheurer) Date: Thu Nov 3 14:41:49 2005 Subject: cvs and project builder Message-ID: <199902041517.QAA00442@xlthlx.sente.ch> On Jan 27, 1999, Steve Loranz wrote: > [...] I think CVL does a great job of translating the cli interface of > cvs into a gui (with the exception of working with trees) I suppose you mean that CVL lacks support for branches. True. We've got a nice idea to do this, but haven't found the time yet. Another important missing feature is support for locks (for instance if you have a policy that wrappers such as nibs should be locked before edited). > [...] It would be great to see CVL provide services for other > apps to utilize It does! In fact it provides a single service, to select, update and display the status of a given file. This should cover 90% of all needs. After activating the service you're in CVL, so you can commit, update, etc. We decided against providing other cvs operations as services, because you usually wants to know the status of a file before doing anything with it. Unfortunately, last time I checked, ProjectBuilder does not put correct file path information on the service pasteboard, so this CVL service is unusable from PB ... > [...] and maybe even better if CVL morphed into a bundle for PB. We have plans not to turn CVL into a bundle but to do a PB bundle that would communicate with CVL. It could be used for stuff like automatically updating file status, or doing all the required cvs operations when you rename a subproject in PB. Marco Scheurer Sen:te From anders at milkweed.com Thu Feb 4 09:35:21 1999 From: anders at milkweed.com (Anders Pytte) Date: Thu Nov 3 14:41:49 2005 Subject: size of YB for PC application binaries Message-ID: Folks, I have an RDR2 application that builds to a size of 1.2 MBs on PPC, but 4.8 MBs on Intel. This is the size of the executable, not including the resources. First of all, why are the binaries so large, even on PPC? The size of the entire project folder (sources, nibs, etc) is only 867 K. But more important, why are the Intel binaries 4 times as large. The application runs fine, but I am obviously doing somethink wrong with my Intel builds. Any information or assistance would be greatly appreciated! Anders. ---------------------------------------------------------------- Anders Pytte Milkweed Software Ferguson Hill Voice: (802) 472-5142 Cabot VT 05647 Internet: anders@milkweed.com ---------------------------------------------------------------- From xavierb at proxyware.fr Thu Feb 4 10:01:30 1999 From: xavierb at proxyware.fr (Xavier Barrier) Date: Thu Nov 3 14:41:49 2005 Subject: OSE/YB for Windows Message-ID: <9902041801.AA00236@proxyware.fr> A non-text attachment was scrubbed... Name: not available Type: text/enriched Size: 1039 bytes Desc: not available Url : /mailman/archive/macosx-dev/attachments/19990204/9f3c919c/attachment.bin From anders at milkweed.com Thu Feb 4 10:15:56 1999 From: anders at milkweed.com (Anders Pytte) Date: Thu Nov 3 14:41:49 2005 Subject: size of YB for PC application binaries In-Reply-To: Message-ID: Thanks to all who immediately responded. I did not know what the "Install" item in the Build Options panel was for! I guess you know where I am in my YB development cycle now! My App on Rhapsody/PPC went from 1.8 MB to 278 KB when built "Install". On YB for W95/NT, the size went from 4.8 MB to 261K. What accounts for this huge difference? Finally, is there a way to build the contents of the resource folder into the executable (I mean on W95/NT only, of course)? My client is concerned that his customers will mistakenly delete the resources folder or move it out of the application folder. Many thanks for your help. Anders. ---------------------------------------------------------------- Anders Pytte Milkweed Software Ferguson Hill Voice: (802) 472-5142 Cabot VT 05647 Internet: anders@milkweed.com ---------------------------------------------------------------- From csaldanh at computeractive.com Thu Feb 4 10:51:29 1999 From: csaldanh at computeractive.com (Chris Saldanha) Date: Thu Nov 3 14:41:49 2005 Subject: size of YB for PC application binaries In-Reply-To: References: Message-ID: <199902041851.NAA00604@saturn.computeractive.com> You wrote: > My App on Rhapsody/PPC went from 1.8 MB to 278 KB when built "Install". On > YB for W95/NT, the size went from 4.8 MB to 261K. What accounts for this > huge difference? The install process strips the binaries of symbol and debugging information which isn't required for running the program. This info is (as you can see!) quite large... --Chris ------------------------------------------------------------------------ Chris Saldanha, Software Analyst | 15 Capella Ct. Suite 128 computerActive Inc. | Nepean, ON, Canada K2E 7X1 csaldanh@computerActive.com (NeXT/MIME)| Voice: (613)225-4824 http://www.computerActive.com/ | FAX: (613)225-1670 ------------------------------------------------------------------------ "The telephone was not invented by Alexander Graham Unitel" -Bell Ad From csaldanh at computeractive.com Thu Feb 4 10:54:13 1999 From: csaldanh at computeractive.com (Chris Saldanha) Date: Thu Nov 3 14:41:49 2005 Subject: OSE/YB for Windows In-Reply-To: <9902041801.AA00236@proxyware.fr> References: <9902041801.AA00236@proxyware.fr> Message-ID: <199902041854.NAA00634@saturn.computeractive.com> You wrote: > I did find some stuff that allows "making an App a service", but the > Serviced server (mine) is not managed correctly. The start/stop commands of > the Service GUI does not get the server app start and stop correctly. > > Does any one ever tried (and succeeded) to get an OSE/YB server app to run > as a Windows NT service? The Windows NT resource kit comes with a program called "srvany.exe" which you can use to make any regular program into an NT service. The resource kit documentation should explain how to set this up. Basically, this srvany program manages the program you wrote and responds correctly to start/stop from the NT Control Panels. --Chris ------------------------------------------------------------------------ Chris Saldanha, Software Analyst | 15 Capella Ct. Suite 128 computerActive Inc. | Nepean, ON, Canada K2E 7X1 csaldanh@computerActive.com (NeXT/MIME)| Voice: (613)225-4824 http://www.computerActive.com/ | FAX: (613)225-1670 ------------------------------------------------------------------------ "The telephone was not invented by Alexander Graham Unitel" -Bell Ad From anders at milkweed.com Thu Feb 4 11:07:24 1999 From: anders at milkweed.com (Anders Pytte) Date: Thu Nov 3 14:41:49 2005 Subject: size of YB for PC application binaries In-Reply-To: <199902041851.NAA00604@saturn.computeractive.com> Message-ID: I wrote: >> My App on Rhapsody/PPC went from 1.8 MB to 278 KB when built "Install". On >> YB for W95/NT, the size went from 4.8 MB to 261K. What accounts for this >> huge difference? Chris wrote: >The install process strips the binaries of symbol and debugging information >which isn't required for running the program. This info is (as you can see!) >quite large... What, then, is the difference between the "app" and "debug" options? Thanks, Anders. ---------------------------------------------------------------- Anders Pytte Milkweed Software Ferguson Hill Voice: (802) 472-5142 Cabot VT 05647 Internet: anders@milkweed.com ---------------------------------------------------------------- From maury at OAAI.COM Thu Feb 4 12:57:18 1999 From: maury at OAAI.COM (Maury Markowitz) Date: Thu Nov 3 14:41:50 2005 Subject: "Infinite" build times - what's wrong? Message-ID: <199902041651.LAA12497@OAAI.COM> I'm running DR2 on a G3/266 with 65M. Any largish build on this machine starts out fine, but then gets slower slower with the drive grinding more and more. After a while a 50 line file takes upwards of a minute to compiler, and eventually it just *stops* (along with the interactivity of the machine). Once this has happened, all builds from that point on _start_ in this state. Quitting/restarting doesn't seem to help, nor does a logout/login. The only "cure" (temporary as it is) seems to be a reboot, which I'm doing upwards of twice a day. Does anyone know exactly what's going on? If this is a memory problem I'll pop the bucks for some more, but I'm not in the mood to spend the money now only to find out it's leaking file handles in the OS or something. Maury From kc at omnigroup.com Thu Feb 4 13:13:57 1999 From: kc at omnigroup.com (Ken Case) Date: Thu Nov 3 14:41:50 2005 Subject: "Infinite" build times - what's wrong? Message-ID: <199902042113.NAA29826@ignem.omnigroup.com> > I'm running DR2 on a G3/266 with 65M. Any largish build on this > machine starts out fine, but then gets slower slower with the > drive grinding more and more. This is caused by a general filesystem I/O bug in DR2. You can work around it by creating and closing Terminal windows until it gets fast again. (I have no idea what the relation is, but it certainly does work. Might take 15-20 windows, though.) Ken From abridge at wheel.dcn.davis.ca.us Thu Feb 4 13:59:15 1999 From: abridge at wheel.dcn.davis.ca.us (Adam Bridge) Date: Thu Nov 3 14:41:50 2005 Subject: creating a subclass of NSFormatter Message-ID: <199902042159.NAA17804@wheel.dcn.davis.ca.us> I'm trying to create my own subclass of NSFormatter. It was all making sense until, during compile-time of course, I got errors related to the method: -(NSString *)stringForObjectValue: (id)anObject When I first saw this I automatically typed (id *)anObject and that's when the compiler chatised me for being naughty. The question is: why am I doing this instead of (id *)anObject which is what everything else in the world seems to use when passing things back and forth. Clearly there's a missing part to my education about the way Objective-C works. Any insight would be most welcome. Thank you Adam Bridge __ Adam Bridge Idea Processing From maury at OAAI.COM Thu Feb 4 14:31:38 1999 From: maury at OAAI.COM (Maury Markowitz) Date: Thu Nov 3 14:41:50 2005 Subject: "Infinite" build times - what's wrong? Message-ID: <199902041825.NAA12615@OAAI.COM> > This is caused by a general filesystem I/O bug in DR2. Ok, thanks for the info you just saved me a SDRAM! > You can work around it by creating and closing Terminal windows > until it gets fast again. (I have no idea what the relation is, but > it certainly does work. Might take 15-20 windows, though.) If I didn't know you better I'd think you were joking. As it is I'll happily do this, and wait for the day that AppleScript is working so I can make the machine do it for me (of course by then I'd hope the bug is fixed). Thanks! Maury From ayoungs at rochester.rr.com Thu Feb 4 15:13:06 1999 From: ayoungs at rochester.rr.com (Andy Youngs) Date: Thu Nov 3 14:41:50 2005 Subject: Yellow Box License Message-ID: Well I found in the WebObject Academic a blurb in Software license agreement to call (512) 919-2645 Apple's Software Licensing. I spoke with a gentlemen (didn't get his name, although he did say he was getting road runner) about Yellow box they supposedly will be taking care of it directly. Has anyone else called ? Regards Andy Youngs From decker at robdecker.com Thu Feb 4 19:59:44 1999 From: decker at robdecker.com (Robert A. Decker) Date: Thu Nov 3 14:41:50 2005 Subject: Stickies on NT/WO 4 Message-ID: I know this is somewhat off-topic and I apologize, but is there anyone else out there that loves being able to use Stickies on NT? I've come to rely on them, just like on my Mac, but they have this annoying bug where notes are duplicated when they're double-clicked in order to collapse them. I'm not sure, but this may happen only when multiple stickies are open. Any potential bug fixes? thanks, rob -- "Robert A. Decker" decker@robdecker.com http://www.robdecker.com/ Programmer Analyst - Health Media Research Lab University of Michigan Comprehensive Cancer Center "Get A Life" quote #3: "I'm not interested in acting. I just have the urge to perform onstage in an animal costume. There's a difference you know." -Chris Elliott From aswift at friday.com Thu Feb 4 20:35:59 1999 From: aswift at friday.com (Adam C Swift) Date: Thu Nov 3 14:41:50 2005 Subject: NSTableView woes In-Reply-To: <199902041518.QAA06580@snaps.frontline-software.dk> References: <199902041518.QAA06580@snaps.frontline-software.dk> Message-ID: <199902050436.WAA02720@redmud.mcs.net> You wrote: > Hi there, > > I have an NSTableView to which I programmatically add columns. My > problem is that the last column always gets a very small width, no > matter what I do with [tableColumn setWidth:]. The total width of the > columns is larger than what the clip view "covers", i.e. the > horizontal scroller is "active". > > Anybody with an idea on how to set the width of the last added column. > > Thx. Geert > I'm not sure if this will help any, but I also ran into trouble setting table view column widths for the last column. I think it happens because of the follow-up actions of the NSTableColumn's setWidth: (marking the table view for display). The table view tries to size either all of the columns or just the last column after any sizing event (window, column, ...). My trivial/hack solution was to write a category to override the NSTableView "sizeLastColumnToFit" method to be a no-op. I decided that the last column should never be resized except programmatically or directly by the user (for better or for worse). Of course, you could make that behavior configurable on a per instance basis by subclassing NSTableView, overriding the method with a conditional invocation of the superclass's implementation and then doing a pose as. - adam From Malcolm_Crawford at plsys.co.uk Fri Feb 5 01:26:29 1999 From: Malcolm_Crawford at plsys.co.uk (mmalcolm crawford) Date: Thu Nov 3 14:41:50 2005 Subject: Stickies on NT/WO 4 In-Reply-To: References: Message-ID: <199902050926.JAA01165@jasmine.plsys.co.uk> Rob wrote: > I know this is somewhat off-topic and I apologize, but is there anyone > else out there that loves being able to use Stickies on NT? > Well, I've had appreciative feedback from a couple of people about Stickies/YB in general. Why me? We actually trained the folks at Apple Cork who wrote it -- it was close to the Friday afternoon project for them. :-) I'm afraid I wasn't desperately happy with breaking UI conventions, and put an extra page into the notes about the importance of sticking to the guidelines, and how easy it was to use the inbuilt classes to ensure this, but I guess that retaining the look and feel of the original Stickies was more important :-) Best wishes, mmalc. From Philippe.Robert at uptime.ch Fri Feb 5 03:22:11 1999 From: Philippe.Robert at uptime.ch (Philippe C.D. Robert) Date: Thu Nov 3 14:41:50 2005 Subject: Mac OS X Server for developers In-Reply-To: References: Message-ID: <9902051122.AA00387@uptime.ch> You wrote: > At 12:47 PM -0600 02/03/1999, Roehl, Thomas wrote: > >I was just curious what people were thinking about Apple's > >announcement that they are going to charge Select developers > >for Mac OS X Server. As far as I know, this is the first time > >that they have charged developers for system software in > >addition to their developer fees. Is this true? Why would they > >do this? > > I have no problem with paying for it. What I have a problem with is not > being able to test my software under it (Mac OS app - NetBoot & BlueBox) > before my clients can buy it and attempt to use it. My clients have the > money to afford to just buy something and assume that my software will work > with it. Read the FAQ, there's an answer for this in it! sweet dreams, Phil --- Philippe C.D. Robert Uptime Object Factory Inc http://www.nice.ch/~phip Software Development http://www.nice.ch/ProjectCenter OpenStep/Unix From btoback at optc.com Fri Feb 5 08:01:31 1999 From: btoback at optc.com (Bruce Toback) Date: Thu Nov 3 14:41:50 2005 Subject: Mac OS X Server for developers Message-ID: <199902051606.JAA30614@landru.optc.com> Scott Anguish writes: > the FAQ posted yesterday say that Premiere and EAP members get it >free of charge. I'm sorry; I didn't see a FAQ. Where is it posted? Thanks, -- 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 herren at flannet.middlebury.edu Fri Feb 5 07:39:11 1999 From: herren at flannet.middlebury.edu (David Herren) Date: Thu Nov 3 14:41:50 2005 Subject: Stickies on NT/WO 4 References: <199902050926.JAA01165@jasmine.plsys.co.uk> Message-ID: Malcolm_Crawford@plsys.co.uk writes: >Well, I've had appreciative feedback from a couple of people about >Stickies/YB in general. It's really funny to see these recent notes about stickies since I just this morning discovered them on NT and I've been very a happy about it since then! I've noticed a bug too in that minimizing all, and then reopening stickies by clicking the button in the task bar, and the stickies lose their close box and grow box...(brief aside: is the task bar the single worst interface convention ever or what...). The best part of stickies for me is that they launch MUCH faster than the Java version of TextEdit, and they allow me to use a yellowbox editor rather than anything else... -- David D. Herren www.cet.middlebury.edu/herren Assoc. Dir. for Tech. & Instruction herren@flannet.middlebury.edu Center for Educational Technology voice: (802)443-5746 5 Court Street, Middlebury, VT 05753 fax: (802)443-2053 From bbum at codefab.com Fri Feb 5 09:53:36 1999 From: bbum at codefab.com (Bill Bumgarner) Date: Thu Nov 3 14:41:50 2005 Subject: NT import/export/extern Was: "Re: gdb techniques?" In-Reply-To: <9902049181.AA918181260@balsa.tiaa-cref.org> Message-ID: All of the "gdb techniques" stuff was very interesting-- but I don't think it'll actually help solve this problem. This problem stinks of the NT ANSI extern's aren't good enough problem. In particular, looking at self via the stack shows a viable object whereas examining self directly shows "". I'd wager that if you did p *self, you would see a structure full of 0/nil? Unfortunately, the NT linker doesn't like the standard "extern" style variable declarations typically used to export symbols from a DLL that may be used by other compiled code (i.e. extern NSString *MyFrameworkException;). Instead, "externs" must be declared differently depending on whether you are compiling the framework itself or using the framework (and its headers) from somewhere else. To do this-- switching the declaration depending on whether you are building or not-- requires a #define that changes depending on whether you are building the framework project or simply using the header from somewhere else. We use a mod to Maekfile.postamble, a FrameworkDefines.template file and add the derived header to three different variables in Makefile.preamble. Details follow. Makefile.preamble: OTHER_PUBLIC_HEADERS = $(DERIVED_SRC_DIR)/FrameworkDefines.h OTHER_PRODUCT_DEPENDS = $(DERIVED_SRC_DIR)/FrameworkDefines.h OTHER_SOURCEFILES = $(DERIVED_SRC_DIR)/FrameworkDefines.h Makefile.postamble additions: $(DERIVED_SRC_DIR)/FrameworkDefines.h: $(DERIVED_SRC_DIR) PB.project Makefile $(SILENT) $(ECHO) Creating FrameworkDefines.h from template. $(SILENT) sed 's//$(NAME)/g' < FrameworkDefines.template > $(DERIVED_SRC_DIR)/FrameworkDefines.h FrameworkDefines.template: /* * DO NOT MODIFY THIS FILE -- IT IS AUTOGENERATED! * * Platform specific defs for externs. This was pretty much * copied from AppKitDefines.h */ // // OpenStep/Mach // #if defined(__MACH__) #ifdef __cplusplus /* * This isnt extern "C" because the compiler will not allow this if it has * seen an extern "Objective-C" */ #define _EXTERN extern #define PRIVATE_EXTERN __private_extern__ #else #define _EXTERN extern #define PRIVATE_EXTERN __private_extern__ #endif #define _EXTERN_IMP PUBLIC_EXTERN #define PRIVATE_EXTERN_IMP PRIVATE_EXTERN // // OpenStep/NT // #elif defined(WIN32) #ifndef BUILDING__DLL #define _WINDOWS_DLL_GOOP __declspec(dllimport) #else #define _WINDOWS_DLL_GOOP __declspec(dllexport) #endif #ifdef __cplusplus #define _EXTERN _WINDOWS_DLL_GOOP extern "C" #define PRIVATE_EXTERN extern "C" #else #define _EXTERN _WINDOWS_DLL_GOOP extern #define PRIVATE_EXTERN extern #endif #define _EXTERN_IMP PUBLIC_EXTERN #define PRIVATE_EXTERN_IMP PRIVATE_EXTERN // // OpenStep/Solaris and/or Solaris/PDO // #elif defined(sun) #ifdef __cplusplus # define _EXTERN extern "C" # define _PRIVATE_EXTERN extern "C" #else # define _EXTERN extern # define _PRIVATE_EXTERN extern #endif #define _EXTERN_IMP _EXTERN #define _PRIVATE_EXTERN_IMP PRIVATE_EXTERN // // PDO HP-UX // #elif defined(hpux) #ifdef __cplusplus # define _EXTERN extern "C" # define _PRIVATE_EXTERN extern "C" #else # define _EXTERN extern # define _PRIVATE_EXTERN extern #endif #define _EXTERN_IMP _EXTERN #define _PRIVATE_EXTERN_IMP PRIVATE_EXTERN #else #error Do not know how to define extern on this platform #endif --- Once that is donw, declare all of your externs as _EXTERN, where is the name of your framework project. It should all just work. Of course, this could be a completely different problem.... b.bum On Thu, 4 Feb 1999, Christopher Nagel wrote: > PLATFORM: NT 4.0 / WO 3.5.1 > > Hello, > > I have an elusive memory/stack-corruption bug that I should very much > like to see exterminated, but I'm afraid my skills with gdb are > insufficient to the task. Does anyone know of resources (or might > teach me) how to really swing in gdb? > > For example, from NeXTanswers, the way cool statement, > > (gdb) po *((id *)($fp +8)) > > does what I thought "po self" should do (but "po self" says " object>"). > > Where can I go to learn what I need to understand in order to know how > to create a command like that? What is $fp, why +8, etc? Anyone know > what information in a DrWatson file would be useful? > > Does anyone have any tried-and-true techniques to share? I remember > Julie Zelinsky (forgive the spelling if it's wrong) gave a very > popular debugging seminar at a NeXTworld a million years ago, but I > never saw any notes from it... > > Thanks very much for any hints/tips/urls you can provide! > > Chris > > > From rfox at neo.rr.com Fri Feb 5 11:15:46 1999 From: rfox at neo.rr.com (rfox) Date: Thu Nov 3 14:41:50 2005 Subject: Mac OS X Server for developers Message-ID: <19990205201846.AAA7066@roadrunner.neo.lrun.com@[24.93.171.160]> >> I have no problem with paying for it. What I have a problem with is not >> being able to test my software under it (Mac OS app - NetBoot & BlueBox) >> before my clients can buy it and attempt to use it. My clients have the >> money to afford to just buy something and assume that my software will work >> with it. Then Phil wrote: > Read the FAQ, there's an answer for this in it! I am sorry. Where is there an answer to his concern in the FAQ? It says developers will get or can buy MOSX when it ships. How does this help his concern above? Robert Fox From brianw at sounds.WA.COM Fri Feb 5 13:52:43 1999 From: brianw at sounds.WA.COM (Brian Willoughby) Date: Thu Nov 3 14:41:50 2005 Subject: NSTableView woes In-Reply-To: <199902041518.QAA06580@snaps.frontline-software.dk> References: <199902041518.QAA06580@snaps.frontline-software.dk> Message-ID: <199902052152.NAA28635@sounds.WA.COM> Geert, [ I have an NSTableView to which I programmatically add columns. My [ problem is that the last column always gets a very small width, [ no matter what I do with [tableColumn setWidth:]. The total width [ of the columns is larger than what the clip view "covers", i.e. [ the horizontal scroller is "active". [ [ Anybody with an idea on how to set the width of the last added [ column. I have several suggestions: It will probably help if at least one of your columns is autosizing and the rest are fixed width. This will allow columns other than the last one to take up the slack in matching the tableView width to the clipView. If none are autosizing, then the last one probably looks awful. You should look into setting the minimum size and maximum size of all columns. When mixing autosizing columns with fixed sized columns, the problem you're experiencing becomes even more complicated. Look at the following code for a category of DBTableVector that I created to help force the size of a column. Updating this from NEXTSTEP 3.3 code to OpenStep/MacOS X Server is left as an exercise for the reader (I believe I have converted it once already, but did not find the code on my first search). // TableVectorsUtil.h #import #import @protocol TableVectorsUtil - (NXCoord)forceSizeTo:(NXCoord)newSize; @end @interface DBTableVector (Util) - (NXCoord)forceSizeTo:(NXCoord)newSize; @end // TableVectorsUtil.m #ifndef MAKE_DEPEND #import "Project.h" #endif #import "TableVectorsUtil.h" @implementation DBTableVector (Util) - (NXCoord)forceSizeTo:(NXCoord)newSize { // WARNING: direct access (read only) of instance variables if (newSize < minSize || 0.0 == minSize) [self setMinSize:newSize]; if (newSize > maxSize || 0.0 == maxSize) [self setMaxSize:newSize]; if (minSize > maxSize) [self setMaxSize:(2 * minSize)]; return [self sizeTo:newSize]; } @end Brian Willoughby Software Design: NEXTSTEP, OpenStep, Rhapsody, MacOS X Sound Consulting Apple Enterprise Alliance Partner NeXTmail welcome Apple is the registered trademark of Apple Computer, Inc. and Apple Records From Philippe.Robert at uptime.ch Sat Feb 6 01:58:24 1999 From: Philippe.Robert at uptime.ch (Philippe C.D. Robert) Date: Thu Nov 3 14:41:50 2005 Subject: Mac OS X Server for developers In-Reply-To: <19990205201846.AAA7066@roadrunner.neo.lrun.com@[24.93.171.160]> References: <19990205201846.AAA7066@roadrunner.neo.lrun.com@[24.93.171.160]> Message-ID: <9902060958.AA00424@uptime.ch> You wrote: > >> I have no problem with paying for it. What I have a problem with is not > >> being able to test my software under it (Mac OS app - NetBoot & BlueBox) > >> before my clients can buy it and attempt to use it. My clients have the > >> money to afford to just buy something and assume that my software will work > >> with it. > Then Phil wrote: > > Read the FAQ, there's an answer for this in it! > > I am sorry. Where is there an answer to his concern in the FAQ? It says > developers will get or can buy MOSX when it ships. How does this help his > concern above? No, no, I replied to another question! You've quoted me wrong...;-) The FAQ just enlightens us about the price of MacOS X Server for developers! sweet dreams, Phil --- Philippe C.D. Robert Uptime Object Factory Inc http://www.nice.ch/~phip Software Development http://www.nice.ch/ProjectCenter OpenStep/Unix From kgb at graduate.net Sat Feb 6 13:48:06 1999 From: kgb at graduate.net (KGB Software) Date: Thu Nov 3 14:41:50 2005 Subject: Java vs ObjC vs C FP performance? Message-ID: Hi, Has anyone compared performance of Java, ObjC, and ANSI C in floating-point intensive apps? Does Java have a significant overhead compared to C? Sorry if this is slightly off-topic. --John Kuszewski From yami at dementian.com Sun Feb 7 00:18:30 1999 From: yami at dementian.com (Steven D. Arnold) Date: Thu Nov 3 14:41:50 2005 Subject: strange errors Message-ID: <000b01be5272$722884c0$3550b5c7@dementian.com> Hi, I am running a program on Yellow Box/NT which compiled fine. When I run it, I get the following errors: (gdb) Starting program: C:/Apple/Projects/calculator/calculator.app/calculator.exe 34020000:C:\Apple\Library\Executables\AppKit.dll 32010000:C:\Apple\Library\Executables\Foundation.dll 31000000:C:\Apple\Library\Executables\nextpdo.dll [failed reading symbols from DLL] "C:\WINNT\system32\MSVCRT.dll": can't read symbols: File format not recognized. 77f00000:C:\WINNT\system32\KERNEL32.dll 77e70000:C:\WINNT\system32\USER32.dll 77ed0000:C:\WINNT\system32\GDI32.dll 77dc0000:C:\WINNT\system32\ADVAPI32.dll 77e10000:C:\WINNT\system32\RPCRT4.dll 70930000:C:\WINNT\system32\SHELL32.dll [failed reading symbols from DLL] "C:\WINNT\system32\SHLWAPI.dll": can't read symbols: File format not recognized. [failed reading symbols from DLL] "C:\WINNT\system32\COMCTL32.dll": can't read symbols: File format not recognized. 776d0000:C:\WINNT\system32\WSOCK32.dll 776b0000:C:\WINNT\system32\WS2_32.dll 776a0000:C:\WINNT\system32\WS2HELP.dll 77fd0000:C:\WINNT\System32\WINMM.dll 77c00000:C:\WINNT\System32\WINSPOOL.DRV 77d80000:C:\WINNT\system32\comdlg32.dll 77b20000:C:\WINNT\system32\ole32.dll 77fc0000:C:\WINNT\System32\mmdrv.dll 6bc00000:C:\WINNT\System32\maestro.dll 728c0000:C:\WINNT\System32\JOYSTICK.DLL 77bf0000:C:\WINNT\System32\rpcltc1.dll 74ff0000:C:\WINNT\System32\rnr20.dll 10000000:C:\Program Files\ToggleMINIMIZE\MinUtil.dll 76ab0000:C:\WINNT\System32\IMM32.dll Program received signal SIGSEGV, Segmentation fault. 0x3100a128 in objc_msgSend () Any idea what might be causing this? It looks like some files are corrupted....perhaps I should reinstall? Thoughts? steve -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman/archive/macosx-dev/attachments/19990207/dd3af128/attachment.html From ayoungs at rochester.rr.com Sun Feb 7 15:36:36 1999 From: ayoungs at rochester.rr.com (Andy Youngs) Date: Thu Nov 3 14:41:50 2005 Subject: Simple Example of pre vs post GUI application Message-ID: Does anyone have a favorite example of an application that was commnad line based that was converted to a GUI one? If not can anyone whip up a quick 5 liner that shows pre GUI code and post GUI code? I can use Interface Builder like a snap and Instantiate and add file to project and all that. Somehow I'm missing that little simple something when it comes to this type of conversion. Thanks in advance Regards Andy From Malcolm_Crawford at plsys.co.uk Sun Feb 7 15:58:17 1999 From: Malcolm_Crawford at plsys.co.uk (mmalcolm crawford) Date: Thu Nov 3 14:41:50 2005 Subject: Simple Example of pre vs post GUI application In-Reply-To: References: Message-ID: <199902072358.XAA04320@jasmine.plsys.co.uk> Andy wrote: > Does anyone have a favorite example of an application that was commnad line > based that was converted to a GUI one? > OpenUp :-) http://www.stepwise.com//Software/OpenUp/ Best wishes, mmalc. From johnt at healthcc.com Sun Feb 7 16:13:02 1999 From: johnt at healthcc.com (John Turnipseed) Date: Thu Nov 3 14:41:50 2005 Subject: Mac OS X Server for developers In-Reply-To: <9902060958.AA00424@uptime.ch> Message-ID: At 4:05 AM -0600 02/06/1999, Philippe C.D. Robert wrote: >No, no, I replied to another question! You've quoted me wrong...;-) No, Phil, you quoted me wrong. I believe your intent was to respond to the original poster's question about Apple charging devlopers for the software. I never asked about the price of Mac OS X to developers. In fact, stated "I have no problem with paying for it." which has the clear implication that I knew that I would be paying for it. Hence the confusion to those who read your post. For the record, here is my original post which you quoted when responding. --------------------- At 12:47 PM -0600 02/03/1999, Roehl, Thomas wrote: >I was just curious what people were thinking about Apple's >announcement that they are going to charge Select developers >for Mac OS X Server. As far as I know, this is the first time >that they have charged developers for system software in >addition to their developer fees. Is this true? Why would they >do this? I have no problem with paying for it. What I have a problem with is not being able to test my software under it (Mac OS app - NetBoot & BlueBox) before my clients can buy it and attempt to use it. My clients have the money to afford to just buy something and assume that my software will work with it. --------------------- >The FAQ just enlightens us about the price of MacOS X Server for developers! Barely. It simply states that Premier and Enterprise partners don't have to pay and that Select will have to pay some unknown amount. If you're enlightened by that, you must be a Premier or Enterprise partner. I guess that makes the cost figure out to be roughly $3000 to Select members until we hear the actual price. Actually $3500, since as of last I checked, there is no "upgrade" policy for ADC memberships. But, since retail was stated by Steve at MacWorld to be $999, I have to believe the cost to Select developers will be much lower than any of that. - Now, not trying to do the ListMoms jobs, but this list is primarily for coding questions as I recall, so I'll attempt to be quiet on this issue from here on out... at least on MOSX-dev. ;^) -=- John _____________________________________________________________________________ John Turnipseed | "If you had to identify, in one word, the reason Mac Driver/Programmer | why the human race has not achieved, and never johnt@healthcc.com | will achieve, its full potential, that word seedman@inetnebr.com | would be: 'MEETINGS.'" -- Dave Barry From amaral at storeria.bio.ou.edu Sun Feb 7 16:40:14 1999 From: amaral at storeria.bio.ou.edu (Jos=?ISO-8859-1?B?6Q==?= Pedro Sousa do Amaral) Date: Thu Nov 3 14:41:50 2005 Subject: Virtual Desktops Message-ID: <1293709417-1997036@storeria.bio.ou.edu> Does Mac OS X Server support virtual desktops similarly to what happens under Linux? I vaguely remember this being discussed a while ago, but at the time I had no clue about how well integrated that was in Linux. I have played with a Mac gizmo that performs a similar function, but I am cutting down on sources of instability. TIA. ZP From paulrs at lgs-systems.com Mon Feb 8 06:21:05 1999 From: paulrs at lgs-systems.com (Paul R. Summermatter) Date: Thu Nov 3 14:41:50 2005 Subject: OSE/YB for Windows In-Reply-To: <9902041801.AA00236@proxyware.fr> References: <9902041801.AA00236@proxyware.fr> Message-ID: <9902081421.AA04195@slab> Xavier, Take a look at Srvany.exe and Srvinst.exe. They come with the NT resource kit and are useful for taking any generic executable and making it into a service. I believe there is another product freely available (ie. you needn't buy the resource kit) called the Invoker, but I don't know where to get this. You might also consider looking at the DOS 'at' command which allows you to run processes at specified times. Regards, Paul You wrote: [ Hi, [ [ I got servers programs 100% OPENSTEP/YB that needs to be running on [ Windows NT like Services. By "Services" i mean the MS implementation of [ crontab into Windows NT. [ [ The only thing i know relating to NT Services, is that such "servers" [ needs to be compile with a re-written main() function call and need to [ be installed into the Services stuff through a dedicated program. [ [ I did find some stuff that allows "making an App a service", but the [ Serviced server (mine) is not managed correctly. The start/stop commands [ of the Service GUI does not get the server app start and stop correctly. [ [ Does any one ever tried (and succeeded) to get an OSE/YB server app to [ run as a Windows NT service? [ [ If yes it could be a point to tell every OSE/YB developers : and my [ self. [ [ --- [ [ Xavier BARRIER [ --------------------------------------------------------- [ ProxyWare SARL 8, rue de l'Isly 75008 Paris France [ Financial Software Company [ [ Tel : (+33) 1 44 69 83 06 [ Fax : (+33) 1 44 69 06 92 [ e-mail : xavierb@proxyware.fr [ --- 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 hunter at lastonepicked.com Mon Feb 8 09:08:08 1999 From: hunter at lastonepicked.com (Hunter Hillegas) Date: Thu Nov 3 14:41:50 2005 Subject: Getting Started with WebObjects Message-ID: <19990208170446274.AAA581@nts1.webhostingprovider.com@firewall.jacobstern.com> How would you folks recommend someone get started with WO? Any good books? Hunter From abridge at wheel.dcn.davis.ca.us Mon Feb 8 09:06:24 1999 From: abridge at wheel.dcn.davis.ca.us (Adam Bridge) Date: Thu Nov 3 14:41:50 2005 Subject: How efficient is NSScanner? Message-ID: <199902081706.JAA10386@wheel.dcn.davis.ca.us> When looking to do field validation for relatively complex entries the use of NSScanner might be nice. Infact the documentation for NSFormatter shows the use of a Scanner object. Was this simply handly or is NSScanner quite efficient? Thanks Adam Bridge __ Adam Bridge Idea Processing From toon at omnigroup.com Mon Feb 8 12:18:45 1999 From: toon at omnigroup.com (Greg Titus) Date: Thu Nov 3 14:41:50 2005 Subject: How efficient is NSScanner? In-Reply-To: <199902081706.JAA10386@wheel.dcn.davis.ca.us> References: <199902081706.JAA10386@wheel.dcn.davis.ca.us> Message-ID: <199902082018.MAA01778@ignem.omnigroup.com> Adam Bridge wrote: > When looking to do field validation for relatively complex entries > the use of NSScanner might be nice. Infact the documentation for > NSFormatter shows the use of a Scanner object. Was this simply >handly or is NSScanner quite efficient? No, NSScanner is not efficient at all (this is more a matter of how the API makes you do things rather than bad implementation). It ends up creating a ton of autoreleased objects. It is very convenient though, so I end up using it quite a bit where speed doesn't matter much. With field validation, since it is a single action in response to the user (who is always very slow as far as the machine is concerned) I think NSScanner would be fine. (What sort of validation are you doing, btw? Just curious...) Where you need to do text processing for bigger things, I highly recommend OFStringScanner in our OmniFoundation framework. The API is different, but it gives you very efficient string scanning, including providing inline C function access to give the compiler plenty of optimization opportunities. It was originally written for OmniWeb's HTML parsing, but it really kicks butt for all sorts of other uses. --Greg --------------------- Greg Titus Omni Development Inc. greg@omnigroup.com From stefan at inis.de Mon Feb 8 12:58:05 1999 From: stefan at inis.de (Stefan Alt) Date: Thu Nov 3 14:41:50 2005 Subject: nib's on NT (NSPopUpList class not loaded) Message-ID: <199902082058.VAA00415@inis-14.inis.de> Hi, I have a programm with nib's created under NeXT. Now I will work under NT. The problem is that if I try to open the nib I get *** class error for 'NSPopUpList' : class not loaded does anybody now this problem and can help stefan From pekeler at codefab.com Mon Feb 8 13:54:07 1999 From: pekeler at codefab.com (Christian Pekeler) Date: Thu Nov 3 14:41:50 2005 Subject: setRefusesFirstResponder:YES and keyboards Message-ID: <9902081654.AA073983@bass> When an NSButton has refusesFirstResponder == YES it won't become first responder when clicked with the mouse. But when I use the button with it's keyEquivalent or mnemonic it becomes first responder. Do I have to do something else besides [button setRefusesFirstResponder:YES] or is this a bug? [WO4.0/WinNT] Thanks, Christian From ngustilo at earthlink.net Tue Feb 9 02:06:05 1999 From: ngustilo at earthlink.net (Nicholas A. Gustilo) Date: Thu Nov 3 14:41:50 2005 Subject: NSStatus Message-ID: <199902082201.OAA19570@grebe.prod.itd.earthlink.net> Hi, I want to use NSStatus class with an application that I am developing but the DR2 documentation is pretty minimal. I have reviewed the header files and cannot figure out how NSStatus (and NSStatusItem) work. If any one has an example they'd be willing to share I'd appreciate it. Thanks in advance. Nick Gustilo. From cmh at greendragon.com Tue Feb 9 12:34:06 1999 From: cmh at greendragon.com (Chris Hanson) Date: Thu Nov 3 14:41:50 2005 Subject: Mac OS X Server device driver development Message-ID: Are any resources available yet on Mac OS X Server device driver development? I'm particularly interested in whether drivers (esp. network drivers) will be portable from BSD to Mac OS X Server, or whether they will need to be rewritten to use DriverKit or the rumored IOKit and just how difficult a task this would be. From gideon at csarc.otago.ac.nz Tue Feb 9 13:58:59 1999 From: gideon at csarc.otago.ac.nz (Gideon King) Date: Thu Nov 3 14:41:50 2005 Subject: Mac OS X Server device driver development In-Reply-To: Message-ID: <9902101058.AA598051@ghnomb> And on this subject, a hardware compatibility guide would be useful - especially for items like RAID systems and backup devices etc. Rumour has it that it is GM now, so you would think that there would be some information on this sort of thing from Apple. It's impossible to make any strategic purchasing plans without any information! Chris Hanson wrote: > Are any resources available yet on Mac OS X Server device driver development? > > I'm particularly interested in whether drivers (esp. network drivers) will > be portable from BSD to Mac OS X Server, or whether they will need to be > rewritten to use DriverKit or the rumored IOKit and just how difficult a > task this would be. > > > ---------------------------------------------------------------- Gideon King | Phone +64-3-479 9018 Technical Manager | Fax +64-3-479 9033 The Black Albatross | University of Otago | Computer Science Applied | e-mail gideon@csarc.otago.ac.nz Research Centre | NeXT mail, MIME ok. PGP key available. Department of Computer Science | P.O. Box 56 | Success always occurs in private, Dunedin | and failure in full view. New Zealand | WWW access: http://www.csarc.otago.ac.nz/~gideon From cdouty at netcom.com Tue Feb 9 14:29:00 1999 From: cdouty at netcom.com (Chris Douty) Date: Thu Nov 3 14:41:50 2005 Subject: Mac OS X Server device driver development In-Reply-To: <9902101058.AA598051@ghnomb> Message-ID: On Tue, 9 Feb 1999, Gideon King wrote: > And on this subject, a hardware compatibility guide would be useful - > especially for items like RAID systems and backup devices etc. Rumour > has it that it is GM now, so you would think that there would be some > information on this sort of thing from Apple. It's impossible to make > any strategic purchasing plans without any information! > > Chris Hanson wrote: > > Are any resources available yet on Mac OS X Server device driver > > development? > > > > I'm particularly interested in whether drivers (esp. network drivers) will > > be portable from BSD to Mac OS X Server, or whether they will need to be > > rewritten to use DriverKit or the rumored IOKit and just how difficult a > > task this would be. Yup, a hardware guide sure would be useful, but I doubt we will get it. Apple seems to release no technical information before a product actually ships. A few days afterwards, stuff starts showing up in TIL. Completely useless for those people who have *any* lead time in their purchasing. Of course, it may be that the hardware list is ridiculously short, e.g. if Apple didn't ship it themselves as part of a support bundle then it is not supported. Now, as to the original question, DR2 had some DriverKit information released and I think that some people managed to get the DriverKit stuff to compile although they had to raid some things (headers) from OS/m 4.2. In any case I would suspect that you have two (OK, three) choices. 1) Write a DriverKit driver if the kit is actually available for the production release of MOSXS. Be prepared to throw away all of this code. 2) Write a Mach Loadable Kernel Server which provide a device interface. Be prepared to throw away all of this code too. 3) Wait patiently, like a good little Koolaid drinker, for daddy Apple to release a beta of IOKit. Of course then all of your code will only work on Mac OS X not the actual operating system you have in your hands. I doubt that any BSD driver will port over cleanly since the kernel interface just aren't there. You can do either DriverKit or LKS. DriverKit is actually a special LKS as I understand it. Information here is derived from past experience and the FAQ at http://developer.apple.com/faq/osxserverfaq.html . As to RAID systems, backup devices, etc. I suspect that they will all just work. If the RAID is completely self-contained and just looks like a big disk to the OS, then it ought to work fine. OS/mach never required stupid FWB SCSI toolkit crap to make a random SCSI disk work. Tape should be pretty similar. What won't work are software RAID or devices like the Adaptec AAA-133 controller. Expect a real short list of approved SCSI controllers and display devices though. -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 aep at apple.com Tue Feb 9 17:12:29 1999 From: aep at apple.com (Andrew Platzer) Date: Thu Nov 3 14:41:50 2005 Subject: NSStatus Message-ID: <199902100112.RAA35244@scv3.apple.com> > Hi, I want to use NSStatus class with an application that I am developing > but the DR2 documentation is pretty minimal. I have reviewed the header > files and cannot figure out how NSStatus (and NSStatusItem) work. If any > one has an example they'd be willing to share I'd appreciate it. Thanks in > advance. Nick Gustilo. It's pretty easy. Let's say you want to just show an icon for a while. Say something like: NSStatusItem* item = [[[NSStatusBar systemStatusBar] statusItemWithLength: NSVariableStatusItemLength] retain]; [item setImage:icon]; It should immediately appear in the system menu bar. Then later on, remove it by saying: [[NSStatusBar systemStatusBar] removeStatusItem:item]; [item release]; You can set an action/target and say [item setHighlightMode:YES] to turn it into a button. You can also set its text, add a menu and enable or disable it. If you want to hide it, set its length to 0 (you can't remove it and add it in later.) Better documentation will be in the next release. Andrew Platzer Application Frameworks Apple Computer, Inc. From jean-michel.cazaux at finindev.com Wed Feb 10 01:00:53 1999 From: jean-michel.cazaux at finindev.com (Jean-Michel Cazaux) Date: Thu Nov 3 14:41:50 2005 Subject: Silent Exceptions ???? Message-ID: <01BE54DC.3F859550@JEAN-MICHEL> We are developping with the YB released with WO 4.0 Beta 1 (release Picasso2F). NSExceptions seems to be in "silent" mode, I mean, in OSE 4.2, when an exception raised, a panel was automaticaly displayed (ut's the way I remember it). Now ***NOTHING*** is visible, no panel, no NSLog. Is there a wy to get former behaviour (using defaults or stuff like this...) or am I missing something ? Thanks in adavance... __________________________________________________________________________________________ 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 martin at rat.se Wed Feb 10 01:34:53 1999 From: martin at rat.se (Martin Wennerberg) Date: Thu Nov 3 14:41:50 2005 Subject: How to get rid of the menu on NT? In-Reply-To: <9902091118.AA17851@next12.Theo-Phys.Uni-Essen.DE> References: <9902091118.AA17851@next12.Theo-Phys.Uni-Essen.DE> Message-ID: <199902100934.KAA29096@baldwin.rat.se> What's the proper way to make a window not have a menu on NT? The NSWindow class always puts a menu at the top of the window and that does not work well in our app. I found one way to do it: Instantiate a NSMenu object in the nib and hook the NSWindow "menu" outlet to it. This gives me a nice menu-less window on NT as well as on OpenStep for mach. However this seems to make the nib-files incompatible between OpenStep for mach 4.2 and WebObjects (NT) 4.0 (due to the NSMenu class having changed it's superclass). The windows need to be able to become main window so I can't use NSPanels. Martin Wennerberg ____________________________________________________ Research & Trade AB Phone: + 46 - 8 - 545 12 265 Fax: + 46 - 8 - 24 54 11 Email: martin@rat.se [MIME & NeXTmail supported] www: http://www.rat.se/homepages/martin Snailmail: Box 7742, S-103 95 Stockholm, SWEDEN Visiting: Kungsgatan 33 From gerard at ina.fr Wed Feb 10 02:16:36 1999 From: gerard at ina.fr (=?iso-8859-1?Q?G=E9rard?= Iglesias) Date: Thu Nov 3 14:41:50 2005 Subject: How to get rid of the menu on NT? References: <199902100934.KAA29096@baldwin.rat.se> Message-ID: <36C15C84.BCBD2115@ina.fr> Martin Wennerberg wrote: > > What's the proper way to make a window not have a menu on NT? The NSWindow > class always puts a menu at the top of the window and that does not work well > in our app. > I use [window setMenu:nil], from NSResponder interface. Sincerely. Gerard From eric at alum.mit.edu Wed Feb 10 02:40:39 1999 From: eric at alum.mit.edu (Eric Hermanson) Date: Thu Nov 3 14:41:50 2005 Subject: How to get rid of the menu on NT? References: <36C15C84.BCBD2115@ina.fr> Message-ID: <36C16227.EB62F5AC@alum.mit.edu> G?rard Iglesias wrote: > Martin Wennerberg wrote: > > > > What's the proper way to make a window not have a menu on NT? The NSWindow > > class always puts a menu at the top of the window and that does not work well > > in our app. > > > > I use [window setMenu:nil], from NSResponder interface. Doing setMenu:nil is a good idea, but if you can live with a panel, you can turn your window into an 'NSPanel' by changing it's class in InterfaceBuilder - by default, panels don't contain menus. Eric From gerard at ina.fr Wed Feb 10 03:31:40 1999 From: gerard at ina.fr (=?iso-8859-1?Q?G=E9rard?= Iglesias) Date: Thu Nov 3 14:41:50 2005 Subject: How to get rid of the menu on NT? References: <36C16227.EB62F5AC@alum.mit.edu> Message-ID: <36C16E1C.F1E627BD@ina.fr> Eric Hermanson wrote: > > G?rard Iglesias wrote: > > > Martin Wennerberg wrote: > > > > > > What's the proper way to make a window not have a menu on NT? The NSWindow > > > class always puts a menu at the top of the window and that does not work well > > > in our app. > > > > > > > I use [window setMenu:nil], from NSResponder interface. > > Doing setMenu:nil is a good idea, but if you can live with a panel, you can turn > your window into an 'NSPanel' by changing it's class in InterfaceBuilder - by > default, panels don't contain menus. > > Eric OK, but if you need your window to be a main window you can't mute it to the NSPanel. Sincerely. Gerard From jocke at rat.se Wed Feb 10 04:01:39 1999 From: jocke at rat.se (Joakim Johansson) Date: Thu Nov 3 14:41:50 2005 Subject: Spurious crashes under YB CR1 [solved] Message-ID: <199902101200.NAA02423@baldwin.rat.se> Several people hav had the same or similar problem as we had, so due to popular demand I'll do a quick recap: Our NEXTSTEP -> YB ported application crashed after a while of normal usage with a backtrace that contained only AppKit classes, notably internal functions of NSMenu: #2 0x3422499f in NSMenu.o:__i_NSMenu__update () The reason for this is that there were off-screen windows with delegates that had been freed which were messaged by NSMenu while it was traversing the windowlist. Now then, the real problem (which still isn't completely pinpointed, but it smells a bit like a potential bug in the AppKit) is that the -window method for a view sometimes returns *nil* even though it should return a proper window. We have a controller in a nib that have an outlet to a view which is in a window (where the controller is the windows delegate). In the -windowWillClose: delegate method, the controller tries to set the windows delegate to nil, by using the often-used construction of [[aView window] setDelegate:nil]. Later, when the window could be freed (we still need it around for a while, but when the controller is released, we're done with it), it would be released. Why [aView window] sometimes returns nil, is still a mystery, it seems to be a race of some sort in the AppKit, but we have other issues to resolve which are of higher priority at the moment. There are plenty of workarounds, use the notifications -object to get the window, use a proper outlet or (as we did for now) use -windowShouldClose: instead. I hope this helps someone else! Best, Joakim -- Joakim Johansson Software Engineer, Research & Trade jocke@rat.se, http://www.rat.se/ From bbum at codefab.com Wed Feb 10 06:06:28 1999 From: bbum at codefab.com (Bill Bumgarner) Date: Thu Nov 3 14:41:50 2005 Subject: Silent Exceptions ???? In-Reply-To: <01BE54DC.3F859550@JEAN-MICHEL> Message-ID: It seems that as OpenStep/NeXTStep have evolved, the default exception handler has become more liberal about letting exceptions ride. I'm not sure I agree with that-- generally, something like an "object doesn't respond to method" exception is a good indication that there is something quite wrong with an object graph and it is likely time to consider death and recovery. I haven't gotten around to implementing a default exception handler that announces unhandled exceptions, but I have put together an exception tracer that announces the exceptions before they are raised. This allows for a couple of modes of debugging that have proven useful. First, it allows one to monitor all of the exceptions that are being raised in an app-- both handled and unhandled exceptions. Monitoring handled exceptions can be quite useful because an exception-- even a handled one-- is indication that the app followed an execution path that involved some kind of an error followed by a recovery. An excessive number of raised/caught/recovered-from exceptions is usually an indication of a performance problem or a bug. Secondly, by providing a known, developer controlled, point of execution guaranteed to be hit before an exception is raised, it becomes much easier to break on and then examine an exception al state before it occurs. Setting a breakpoint on an exception handler is generally useless-- by that time the stack has been wiped of all the frames down to where the exception was actually raised. By catching it before, one can actually step up the stack and examine what was really going on. It would not be hard to extend the code to provide easy breakable hooks that are only hit on specific exceptions. Enclosed is the subclass of NSException that we use-- it poses as NSException and overrides -raise. It comes from the Mssh project, so to see the implementation of anything it might depend on, see the mssh project in ftp://ftp.codefab.com/pub/unsupported/. b.bum /* Copyright 1999 by CodeFab, Inc. All rights reserved. This notice may not be removed from this source code. The use and distribution of this software is governed by the terms of the MiscKit license agreement. Refer to the license document included with the MiscKit distribution for the terms. Details of the license agreement may be found at www.misckit.com or within the MiscKit distribution. There may be more information about this software on www.codefab.com. Send queries to ask_codefab@codefab.com. */ // import APIs to NeXT provided frameworks #import // import local framework's API #import "CodeFabSupport.h" // import API required througout this file's scope #import "CFRExceptionTracer.h" /*" * Exception raised when one of the assertion macros like CFRValidateMethodArgument() fails. "*/ NSString *CFRInvalidVariableException = @"CodeFabRuntime: Invalid Variable Value Exception."; @implementation CFRExceptionTracer /*" * A useful debugging tool that provides exception tracing. When the exception tracer is installed and enabled, a message will * be logged on the console whenever the #{-raise} method of #NSException is about to be invoked. As such, this class allows the * developer to also monitor exceptions that are handled within an NS_HANDLER block. "*/ static BOOL exceptionTracingInstalled = NO; static BOOL exceptionTracingEnabled = NO; + (void) testExceptionTracing /*" * Tests the exception tracing functionality provided by this class. Of course, this method invokes * #{+installExceptionTracing} and, as such, will modify the runtime in a manner that cannot be undone. "*/ { NSAutoreleasePool *pool; id randomObject; pool = [[NSAutoreleasePool alloc] init]; randomObject = [[NSObject alloc] init]; [self installExceptionTracing]; NS_DURING [self logMethod: _cmd withFormat: @"Raising...."]; [(NSArray *) randomObject objectAtIndex: 0]; [self logMethod: _cmd withFormat: @"SHOULDN'T HAVE HAPPENED (1)"]; NS_HANDLER [self logMethod: _cmd withFormat: @"Handled."]; [pool release]; NS_VOIDRETURN; NS_ENDHANDLER [self logMethod: _cmd withFormat: @"HOULDN'T HAVE HAPPENED (2)"]; [pool release]; } + (void) installExceptionTracing /*" * Installs the exception tracing mechanism. Once installed, the tracer cannot be uninstalled. The installation is * performed by telling the #CFRExceptionTracer class to pose as the NSException class. * * Internally, this method creates and destroys an instance of NSAutoreleasePool around the installation of the tracer. As * such, this method can be invoked at any time during application start up. * * Upon successful invocation, the exception tracer will be installed and enabled. "*/ { NSAutoreleasePool *pool; if (exceptionTracingInstalled) return; pool = [[NSAutoreleasePool alloc] init]; [self logMethod: _cmd withFormat: @"Installing exception tracing..."]; [CFRExceptionTracer class]; // ensure classes are intialized [CFRExceptionTracer poseAsClass: [NSException class]]; [pool release]; exceptionTracingInstalled =YES; exceptionTracingEnabled = YES; } + (void) setExceptionTracingEnabled: (BOOL) aFlag /*" * Enables and disables the exception tracer. Enabling the tracer does not automatically install it. "*/ { exceptionTracingEnabled = aFlag; } + (BOOL) exceptionTracingEnabled /*" * Returns YES if the exception tracer is enabled (does not test to see if it is installed). "*/ { return exceptionTracingEnabled; } + (BOOL) exceptionTracingInstalled /*" * Returns YES if the exception tracing mechanism has been installed. "*/ { return exceptionTracingInstalled; } - (volatile void)raise /*" * Upon installation, the exception tracer class #CFRExceptionTracer poses as #NSException. As such, this method overrides * #NSException's implementation of #{-raise}. If exception tracing is enabled, this method logs a nicely formatted message * via #{-logMethod:withFormat:} indicating the name, reason, and userInfo for the exception that is about to be raised. * * Finally, this method invokes #NSException's original implementation of #{-raise} to cause the normal exception behaviour to * occur. "*/ { if (exceptionTracingEnabled) [self logMethod: _cmd withFormat: @"Will raise exception:\n\tName: %@\n\tReason: %@\n\tuserInfo: %@\n", [self name], [self reason], [self userInfo]]; [super raise]; } @end From paulrs at lgs-systems.com Wed Feb 10 04:24:49 1999 From: paulrs at lgs-systems.com (Paul R. Summermatter) Date: Thu Nov 3 14:41:50 2005 Subject: Spurious crashes under YB CR1 [solved] In-Reply-To: <199902101200.NAA02423@baldwin.rat.se> References: <199902101200.NAA02423@baldwin.rat.se> Message-ID: <9902101224.AA06095@slab> Joakim wrote: [ [ [ Now then, the real problem (which still isn't [ completely pinpointed, but it smells a bit like [ a potential bug in the AppKit) is that the [ -window method for a view sometimes returns *nil* [ even though it should return a proper window. [ [ Joakim, I don't understand this. There is no reason why a view should always return a value when asking for its window. Take the following as an example: [mySwapViews addObject:aSwapView]; [aSwapView removeFromSuperview]; At this point, aSwapView is still a perfectly valid view which can be put into a window at some later point in time, but if you ask the view for its window it will return nil. A similar situation may arise if you are doing the following: myOutletToASwapView = [aSwapView retain]; In this case, aSwapView's window may be freed by someone else causing it to be removed from the view heirarchy in the window, but because you have retained it, it will hang around, and would return a nil value when asked for a window. In sum, there are plenty of perfectly reasonable situations under which a view will return nil when asked for its window. 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 jocke at rat.se Wed Feb 10 09:49:46 1999 From: jocke at rat.se (Joakim Johansson) Date: Thu Nov 3 14:41:50 2005 Subject: Spurious crashes under YB CR1 [solved] Message-ID: <199902101748.SAA09956@baldwin.rat.se> Just to stall other's with the same (reasonable) argument... (I didn't see that Paul CC:ed the list, thus the post of my reply mail to him) Joakim From: Joakim Johansson Date: Wed, 10 Feb 99 16:55:19 +0100 To: paulrs@lgs-systems.com Subject: Re: Spurious crashes under YB CR1 [solved] In-Reply-To: <199902101200.NAA02423@baldwin.rat.se> Hi Paul, > I don't understand this. There is no reason why a view should > always return a value when asking for its window. Take the following as an > example: > [snip> > In sum, there are plenty of perfectly reasonable situations under > which a view will return nil when asked for its window. I guess I was a bit unclear, this view is static in the window and is never removed (it's a spreadsheet-like view that is completely tied to the window/controller). We've got inspector windows where we are swapping views in/out from the window, but this is not one of them. (thus we believe it was a bit unreasonable for it to return nil just once in a while...) Cheers, Joakim -- Joakim Johansson Software Engineer, Research & Trade jocke@rat.se, http://www.rat.se/ From geek at geeksrus.com Wed Feb 10 10:14:46 1999 From: geek at geeksrus.com (Steven W. Riggins) Date: Thu Nov 3 14:41:50 2005 Subject: Drives with OSXS Message-ID: <199902101817.KAA30366@disneyland.geeksrus.com> Can OSXS work on a blue G3 internal ATA drive or does it only work on the SCSI drive? Steve From geek at geeksrus.com Wed Feb 10 10:16:12 1999 From: geek at geeksrus.com (Steven W. Riggins) Date: Thu Nov 3 14:41:50 2005 Subject: Drives with OSXS Message-ID: <199902101819.KAA30374@disneyland.geeksrus.com> Oops sorry off by one in my mailing list folder, bad programmer bad! :) Steve From fabienr at ncmi.com Wed Feb 10 11:12:39 1999 From: fabienr at ncmi.com (Fabien Roy) Date: Thu Nov 3 14:41:50 2005 Subject: Nil object returning NaN for a float method In-Reply-To: <9902101224.AA06095@slab> References: <9902101224.AA06095@slab> Message-ID: <199902101913.LAA04813@ignem.omnigroup.com> Hi, I just tested some code on nil object, here is a summary of the results: It seems that (maybe as a side effect) the ObjC runtime of Apple/NeXT is returning NaN for the float method of a nil object. Tested on: NEXSTEP 3.3 Intel OPENSTEP 4.2 Mach Intel Openstep 4.2 NT WO 3.51 NT WO 4.0 NT WO 4.0 Deployment Solaris 2.6 Could somebody test it also on DR[12]. Here is the code: /************************** cut here *******************************/ #import @interface MyNSObject:NSObject { } -(float) myFloatValue; -(double) myDoubleValue; -(NSString *) myNSStringValue; @end @implementation MyNSObject -(float) myFloatValue { return 1.11111; } -(double) myDoubleValue { return (double)2.2222222222222; } -(NSString *) myNSStringValue { return @"A string"; } @end int main(int argc, const char *argv[]) { MyNSObject *myObject=nil; float myFloat; double myDouble; /* here to examine with gdb */ myFloat = [myObject myFloatValue]; myDouble = [myObject myDoubleValue]; NSLog(@"\nmyObject: %@ ,\n[(MyNSObject *)0x0 myFloatValue] = " @"%f,\n[(MyNSObject *)0x0 myDoubleValue] = %f,\n[myObject " @"myNSStringValue] = %@\n\n", (MyNSObject *)0x0, [(MyNSObject *)0x0 myFloatValue], [(MyNSObject *)0x0 myDoubleValue], [(MyNSObject *)0x0 myNSStringValue]); exit(0); } /************************** cut here *******************************/ Later. Fabien --- Fabien L Roy NationsBanc Montgomery Securities, LLC voice: (704)386-75-76 100 North Tryon Street fax: (704)388-95-64 NC1-007-09-08 Charlotte NC 28255 Beeper: 143-9722 (1-800-946-46-46) or http://www.MobileComm.com/message/ email: fabienr@ncmi.com (NeXT/Mime) Pager-email 1439722@mobilecomm.net From ckane at apple.com Wed Feb 10 11:28:51 1999 From: ckane at apple.com (Chris Kane) Date: Thu Nov 3 14:41:50 2005 Subject: Nil object returning NaN for a float method Message-ID: <199902101928.LAA19588@scv1.apple.com> Fabien Roy writes: > I just tested some code on nil object, here is a > summary of the results: It seems that (maybe as a > side effect) the ObjC runtime of Apple/NeXT is > returning NaN for the float method of a nil object. This is allowed by the ObjC language. The result of a message expression with a nil receiver expression is only defined where the return type of the method is "id" or a pointer to an explicitly typed object ('object' in the object-oriented, not ANSI C, sense), and is undefined in all other cases, including other pointer types. In your case, the garbage in the float-return register or stack area when the message call returns happens to look like NaN. Chris Kane Apple Computer, Inc. From fabienr at ncmi.com Wed Feb 10 11:49:05 1999 From: fabienr at ncmi.com (Fabien Roy) Date: Thu Nov 3 14:41:50 2005 Subject: Nil object returning NaN for a float method In-Reply-To: <199902101928.LAA19588@scv1.apple.com> References: <199902101928.LAA19588@scv1.apple.com> Message-ID: <199902101949.LAA13165@ignem.omnigroup.com> You wrote: > This is allowed by the ObjC language. The result of a message > expression with a nil receiver expression is only defined where the > return type of the method is "id" or a pointer to an explicitly > typed object ('object' in the object-oriented, not ANSI C, sense), > and is undefined in all other cases, including other pointer types. > In your case, the garbage in the float-return register or stack > area when the message call returns happens to look like NaN. > > Chris Kane > Apple Computer, Inc. > It was for curiosity. I wanted to know if this behavior is consistent across architectures. Thanks. Fabien. --- Fabien L Roy NationsBanc Montgomery Securities, LLC voice: (704)386-75-76 100 North Tryon Street fax: (704)388-95-64 NC1-007-09-08 Charlotte NC 28255 Beeper: 143-9722 (1-800-946-46-46) or http://www.MobileComm.com/message/ email: fabienr@ncmi.com (NeXT/Mime) Pager-email 1439722@mobilecomm.net From paulrs at lgs-systems.com Wed Feb 10 13:54:09 1999 From: paulrs at lgs-systems.com (Paul R. Summermatter) Date: Thu Nov 3 14:41:50 2005 Subject: Modal Loops & Exception handlers Message-ID: <9902102154.AA06471@slab> Hey guys, We're having a problem and are trying to better understand all of the issues involved and would appreciate any info on this topic. I have, in the past, written code which successfully starts and runs a panel in a modal loop which shows progress and gives the user an opportunity to cancel that progress. I needed to accomplish almost the same thing in another part of our app, so I tried to reuse the bulk of the code that I knew worked. Essentially, the code is of the form aSession = [NSApp beginModalSessionForWindow:myProgressPanel]; while ([NSApp runModalSession:aSession] == NSRunContinuesResponse) { // do something } [NSApp endModalSession:aSession]; However, in this new object, the code is failing to work. Sometime after I call beginModalSession but before runModalSession: a message appears in my debugger that says: 'Exception handlers were not removed properly'. Just before I get to the line of code that calls endModalSession, another message appears in gdb that says: 'Attempt to remove unrecognized exception handler'. Then, when endModalSession is called, the app crashes with a memory access exception in runModalCleanup (). I assumed that these messages had to do with an exception handler of the form NS_DURING... somewhere in my code. Right before begining the modal session, I set a break point and walked up the stack to verify that no where in my calling code is there any exception handling block of any form. We do have an exception handler in our Main, but this has never caused us a problem before with running modal sessions. Does anyone know what might be going on here? Also, can anyone clarify what the NSApp documentation for beginModalSession means when it says: "Do not use NS_DURING constructs to send an endModalSession: message in the event of an exception."? An example of exactly what NOT to do would be greatly appreciated. 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 holger at wizards.de Wed Feb 10 15:10:48 1999 From: holger at wizards.de (Holger Hoffstaette) Date: Thu Nov 3 14:41:50 2005 Subject: WO4: PB method browser still broken? Message-ID: <199902102310.AAA13692@ragnarok.en.uunet.de> [I originally sent this to the webobjects list only, although people using YB 'only' might find this helpful as well. Holger] Some days ago I wrote: >is it only me or is the WO4/NT ProjectBuilder still not able to correctly >browse source files? Turns out that this behavior is really still part of WO 4.0/NT as well. Amazing. :-( Out of sheer hate for unnecessary bugs like this I decided to let the Objective-C runtime speak for me. The result is a little PB bundle that fixes the bug on both WO 3.5.1/Mach and WO 4/NT. It also allows you to modify the number of columns in the WO 3.5.1 PB code browser, although that feature is a little unreliable and turned off by default. Use the source. You can get the bundle here: http://www.wizards.de/holger/warez.html have fun & let me know if you have problems or improve it. Regards, Holger (who is *really* happy with the new WO4/YB-NT otherwise..good work!) From holger at wizards.de Thu Feb 11 01:54:31 1999 From: holger at wizards.de (Holger Hoffstaette) Date: Thu Nov 3 14:41:50 2005 Subject: WO4: PB method browser still broken? Message-ID: <199902110954.KAA16422@ragnarok.en.uunet.de> Hi, Several people noted that they cannot get the bundle to link correctly. I forgot to mention that you have to add -undefined suppress to the Linker Options if you're using NT. IIRC that's a recently discussed bug in the NT makefiles. Holger From paulrs at lgs-systems.com Thu Feb 11 12:17:18 1999 From: paulrs at lgs-systems.com (Paul R. Summermatter) Date: Thu Nov 3 14:41:50 2005 Subject: Modal Loops & Exception handlers Followup Message-ID: <9902112017.AA07094@slab> Hey guys, I didn't hear even a peep after my post on this topic, so I guess the problem must be fairly esoteric. In any event, my partner and I built two separate examples of this issue and came up with the same results. For whatever reason, notifications can throw a wrench into the process of creating a modal loop. In particular, if you have a non-ui object which throws up a notification requesting that a UI object present UI to show progress and you use subsequent notifications to send updated information about the progress, your app ends up crashing in the manner which I described the other day in my post. However, if instead of notifications, you use some direct messaging scheme, everything seems to just work. My question still is: WHY? I now know how to avoid this problem, but I understand it no better than before. 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 sanguish at digifix.com Thu Feb 11 22:46:11 1999 From: sanguish at digifix.com (Scott Anguish) Date: Thu Nov 3 14:41:50 2005 Subject: NSDocument/NSDocumentController questions Message-ID: <199902120646.BAA09636@digifix.com> We have a 'not too uncommon' situation that we're having to work though.. We use a wrapper approach for our files, since many of the 'sub-files' for our application can be huge (images, movies etc..) Our 'New document' situation is as a result a little different. When you do a new, you're presented with the Save as... dialog, and then if you successfully save, you're shown the new empty document. So we need to be able to prevent the "New Document" by default behaviour and do it this way instead.. --- The other problem, and perhaps the more important one.. is this When we do a save, NSDocument moves the old file (in this case a directory) to a new location, and then gives us the new path to save to... The problem with this behaviour is that we don't get given that temporary path location, so we don't know where the original files are to copy them to the new path.. For stuff like NSImage etc... that handle the containment of the files for us, thats not a major problem.. we can just write them out.. However for Quicktime movies that can top out at 80 Megs easy.. and the fact that there isn't a self-contained object for it, we're stymied... What are the options here? Ideally NSDocument should save the new file to a renamed location, and then move that to the old file name when its done.. that way were given the new file location, and can look at the Document filename for the original location.. That would fix the problem.. or at least prevent it.. From adhamh at mail.utexas.edu Thu Feb 11 23:00:41 1999 From: adhamh at mail.utexas.edu (adhamh) Date: Thu Nov 3 14:41:50 2005 Subject: No subject Message-ID: <199902120657.WAA29828@scv3.apple.com> I'm a novice c programmer who has made it through day 9 of "Learn C in 21 days" and I trying to port the cable mode client for linux to Mac OS X Server. The source is located at www.vortech.net/rrlinux. These apps build on FreeBSD and Redhat linux, so it seemed a good bet that the port would not extermely difficult, but I may be in over my head. The package is made up of three apps rrdhcpd, rrlogin, and rrconf. I have had success building rrlogin (with errors) and rrconf (worked the first time), but rrdhcp will not build. I have tired both the 1.0.3 and 2.0 versions, with similar errors. Below is the output of the make command for the 1.0.3 version, the debug flag was not set. Similar errors occur with the 2.0 version but they originate in the inet.h header. Can anyone suggest a solution, or do I need to give more info? Thanks! adhamh@mail.utexas.edu cc -g -Wall -O2 -c main.c In file included from main.c:32: /System/Library/Frameworks/System.framework/Headers/bsd/net/if.h:110: warning: `struct sockaddr' declared inside parameter list /System/Library/Frameworks/System.framework/Headers/bsd/net/if.h:110: warning: its scope is only this definition or declaration, /System/Library/Frameworks/System.framework/Headers/bsd/net/if.h:110: warning: which is probably not what you want. /System/Library/Frameworks/System.framework/Headers/bsd/net/if.h:286: field `ifru_addr' has incomplete type /System/Library/Frameworks/System.framework/Headers/bsd/net/if.h:287: field `ifru_dstaddr' has incomplete type /System/Library/Frameworks/System.framework/Headers/bsd/net/if.h:288: field `ifru_broadaddr' has incomplete type /System/Library/Frameworks/System.framework/Headers/bsd/net/if.h:309: field `ifra_addr' has incomplete type /System/Library/Frameworks/System.framework/Headers/bsd/net/if.h:310: field `ifra_broadaddr' has incomplete type /System/Library/Frameworks/System.framework/Headers/bsd/net/if.h:311: field `ifra_mask' has incomplete type In file included from /System/Library/Frameworks/System.framework/ Headers/bsd/net/if.h:340, from main.c:32: /System/Library/Frameworks/System.framework/Headers/bsd/net/if_arp.h:55: field `arp_pa' has incomplete type /System/Library/Frameworks/System.framework/Headers/bsd/net/if_arp.h:56: field `arp_ha' has incomplete type In file included from main.c:33: /System/Library/Frameworks/System.framework/Headers/bsd/arpa/inet.h:45: warning: `struct in_addr' declared inside parameter list /System/Library/Frameworks/System.framework/Headers/bsd/arpa/inet.h:46: warning: `struct in_addr' declared inside parameter list /System/Library/Frameworks/System.framework/Headers/bsd/arpa/inet.h:46: warning: parameter has incomplete type /System/Library/Frameworks/System.framework/Headers/bsd/arpa/inet.h:48: warning: parameter has incomplete type /System/Library/Frameworks/System.framework/Headers/bsd/arpa/inet.h:50: warning: parameter has incomplete type In file included from main.c:39: client.h:144: warning: `struct ifinfo' declared inside parameter list client.h:145: warning: `struct ifinfo' declared inside parameter list main.c: In function `main': main.c:150: warning: implicit declaration of function `ifReset' make: *** [main.o] Error 1 From yonas at powertrader.com Sat Feb 13 17:58:05 1999 From: yonas at powertrader.com (yonasj) Date: Thu Nov 3 14:41:50 2005 Subject: Solaris/Intel and PDO. Message-ID: <36C62DAD.66106905@powertrader.com> Does anybody know if PDO will work for Solaris/Intel. We are considering this due to the lack of MOX/Intel. -- Thanks. -- Yonas. From adhamh at mail.utexas.edu Sun Feb 14 09:24:09 1999 From: adhamh at mail.utexas.edu (adhamh) Date: Thu Nov 3 14:41:50 2005 Subject: dhcp client for nextstep/openstep/mxs?? (cable modem) Message-ID: <199902141720.JAA11438@scv3.apple.com> I'm trying to get a cable modem to work with MXS (Mac OS X Server), I have compiled a the login program successfully, but I'm having trouble getting the dhcp client to compile. Does anyone know of dhcp client that works on nextstep or openstep that would be easy to port, or that may just work on MXS? From Olivier.Saut at crans.ens-cachan.fr Mon Feb 15 01:11:59 1999 From: Olivier.Saut at crans.ens-cachan.fr (Olivier Saut) Date: Thu Nov 3 14:41:50 2005 Subject: dhcp client for nextstep/openstep/mxs?? (cable modem) In-Reply-To: <199902141720.JAA11438@scv3.apple.com> Message-ID: <9902151011.AA591582@marid> adhamh@mail.utexas.edu (adhamh) wrote: > I'm trying to get a cable modem to work with MXS (Mac OS X Server), I > have compiled a the login program successfully, but I'm having trouble > getting the dhcp client to compile. Does anyone know of dhcp client > that works on nextstep or openstep that would be easy to port, or that > may just work on MXS? > > I have also tried to use dhcp client on Rhapsody. I downloaded dhcp2.0b1pl6 on the dhcp site. The client compiles well, the makefile recognizes Rhapsody. But I did not find a script that works. My network uses static IPs so I did not go further. Feel free to e-mail me if you want details. -- Olivier From mmiller at iol.ie Mon Feb 15 02:52:16 1999 From: mmiller at iol.ie (Mark Miller) Date: Thu Nov 3 14:41:50 2005 Subject: _NULL_THUNK_DATA Message-ID: I am having trouble compiling my version of the OmniNetworking framework on windows. I get a link error, LNK2001 (ok dave), unresolved external symbol "AvatarOmniBase_NULL_THUNK_DATA". What is NULL_THUNK_DATA when it is at home? I have renamed the frameworks because I made some small changes. They compile and run fine on Mac OS X Server. Thanks From: Mark Miller, CTO -- AVATAR Works Limited, 154 Leinster Road, Dublin 6, IRELAND -- Phone : +353-1-676 6099 (from US 011-353-1-676 6099) Fax : +353-1-676 6401 (from US 011-353-1-676 6401) Email : mailto:mark.miller@avatarworks.com Web : http://www.avatarworks.com/ -------------- AVATAR - people friendly groupware ------------- From csaldanh at computeractive.com Mon Feb 15 07:19:37 1999 From: csaldanh at computeractive.com (Chris Saldanha) Date: Thu Nov 3 14:41:50 2005 Subject: Solaris/Intel and PDO. In-Reply-To: <36C62DAD.66106905@powertrader.com> References: <36C62DAD.66106905@powertrader.com> Message-ID: <199902151519.KAA26811@saturn.computeractive.com> You wrote: > Does anybody know if PDO will work for Solaris/Intel. > > We are considering this due to the lack of MOX/Intel. Apple does not ship anything for Solaris/Intel, only Solaris/SPARC. --Chris ------------------------------------------------------------------------ Chris Saldanha, Software Analyst | 15 Capella Ct. Suite 128 computerActive Inc. | Nepean, ON, Canada K2E 7X1 csaldanh@computerActive.com (NeXT/MIME)| Voice: (613)225-4824 http://www.computerActive.com/ | FAX: (613)225-1670 ------------------------------------------------------------------------ "The telephone was not invented by Alexander Graham Unitel" -Bell Ad From maury at OAAI.COM Mon Feb 15 13:50:09 1999 From: maury at OAAI.COM (Maury Markowitz) Date: Thu Nov 3 14:41:50 2005 Subject: EPS's vs. TIFF's clipping to userPath's? Message-ID: <199902152145.QAA04729@OAAI.COM> I'm filling shapes with images and having a problem that doesn't seem to make any sense. Basically the shape has a userPath, a Stroke (custom object) and a Fill (ditto), and in the shape's draw it does a... PSgsave(); [path send]; PSclip(); [fill draw]; PSgrestore(); Simple enough. The fills come in a number of varieties, the image fill simply keeps an NSImage and does a... [[image bestRepresentationForDevice:nil] draw]; In it's draw. The problem is that TIFF's work perfectly, but EPS's don't seem to clip. Is this a known issue? It _seemed_ to work under 4.2, but that was a long time ago and I'm not utterly convinced it worked - but I'm pretty close. Maury From maury at OAAI.COM Mon Feb 15 14:04:04 1999 From: maury at OAAI.COM (Maury Markowitz) Date: Thu Nov 3 14:41:50 2005 Subject: window size reporting problem? Message-ID: <199902152159.QAA04748@OAAI.COM> Here's another... I've added some code to make the window size to the max needed for the document inside, I like the resulting interface. During the DR2->4.2 porting process though the porter noted that it didn't seem to work right under DR2/Intel. This was odd, as it worked fine on my G3. A little experimenting showed that it wasn't actually working right at all. A little more debugging suggests there's a bug in NSWindow somewhere. Basically our window has two main views, an info area at the top of the window, and the main scrollView below it. To get the size of the small view at the top, I do this... contentRect = [[[self window] contentView] bounds]; scrollRect = [scrollView frame]; Subtracting the two gives me the height of the bar at the top, and anything else we add in the future. From there I work my way out from the document size (typically 8.5x11") add in those values, then figure out how big the window should be. Presto. Here's the problem, when I put the window into the background and then bring it to the front, contentRect ends up with a bad value. Specifically the pre-covering of the window gets... (gdb) print contentRect $7 = { origin = { x = 0, y = 0 }, size = { width = 521, height = 717 } } Looks good to me. But then after covering the window (even partially) and bringing it back to the front, I get... (gdb) print contentRect $8 = { origin = { x = 178.828613, y = 0 }, size = { width = -1.999403, height = 0 } } *blink*. The window IS in the front, I click on it to bring it to the front, and then use the resize corner which drops me into this code. All the other values are fine, both the "needed" view size and the scrollRect are valid. Any ideas? Maury From daryl at montagetech.com Mon Feb 15 17:04:09 1999 From: daryl at montagetech.com (Daryl Thachuk) Date: Thu Nov 3 14:41:50 2005 Subject: IB connection inspector on WOF4/NT Message-ID: <199902160104.SAA00374@cadvision.com> Hi dee Ho Can someone confirm that the connection inspector for IB running on WOF4/NT is working? Whenever I examine an AppKit object connection in IB I get a blank panel. No connection inspector, no message, nothing. The only time I can get the connection inspector is when I inspect the First Responder object. Everything else comes up blank. -daryl --- Daryl Thachuk daryl@montagetech.com Montage Technologies Inc. http://www.montagetech.com From derek at csarc.otago.ac.nz Mon Feb 15 19:59:17 1999 From: derek at csarc.otago.ac.nz (Derek Rendall) Date: Thu Nov 3 14:41:50 2005 Subject: Java EOF tool problems Message-ID: <36C8ED15.A2727EBC@csarc.otago.ac.nz> Hi, I am writing a java EOF tool that takes a text file and inserts records into the database. When I created the tool, it gave me an Objective C main file (I could not see a way to change it to Java). The rest of my code is in Java, as my EOF framework is in Java. My first problem was trying to call my java code from the main (OC) file. After looking at previous postings I implemented the following: NSJavaVirtualMachine *myVM = [NSJavaVirtualMachine defaultVirtualMachine]; Class myJavaClass = [myVM findClass:@"myClass"]; id myJavaObject = [[myJavaClass alloc] init]; This appeared to work - I could call my class. So, the next step was to do some EOF stuff in my class. As soon as I try to fetch (both EOUtilities.objectsWithQualifierFormat() and ec.objectsWithFetchSpecification()) I get the following error: NSInternalInconsistencyException: NSJavaVirtualMachine: multiple virtual machines are not supported in a single process I have checked, and my original call to [NSJavaVirtualMachine defaultVirtualMachine] is creating a new one. I thought it would use that as the default one - what gives? Is EOF creating it's own defaultVM? Most importantly - how do I get around this? Thanks -- Derek Rendall Senior Software Engineer Black Albatross University of Otago P.O. Box 56 Dunedin New Zealand ph +64 3 4795468 fax +64 3 4799033 http://www.albatross.co.nz From kcd at jumpgate.com Mon Feb 15 23:36:25 1999 From: kcd at jumpgate.com (Kenneth C. Dyke) Date: Thu Nov 3 14:41:50 2005 Subject: Thread Local Storage? Message-ID: <199902160736.XAA03969@babylon5.jumpgate.com> Is there a really fast way to do thread-local storage under Mach? The only way I can think of would be to use thread_self() with a hash table that points to the local data you need. Unfortunately I think this could be far too slow for my eventual needs, since I *think* thread_self() is a system call. What would be "cool" would be if you could set aside some set of memory pages at a fixed address within my task's address space, but have those pages be different for every thread, which I don't think Mach supports. Any ideas would be great. Unfortunately I am stuck dealing with an API that has been around for a while, so I can't do something like pass the state information in to each function. ;) -Ken From bungi at omnigroup.com Tue Feb 16 01:03:34 1999 From: bungi at omnigroup.com (Timothy J. Wood) Date: Thu Nov 3 14:41:50 2005 Subject: Thread Local Storage? Message-ID: <199902160903.BAA11799@mungi.omnigroup.com> A non-text attachment was scrubbed... Name: not available Type: text/enriched Size: 4394 bytes Desc: not available Url : /mailman/archive/macosx-dev/attachments/19990216/b210d9c2/attachment.bin From marcel at system.de Tue Feb 16 00:45:38 1999 From: marcel at system.de (Marcel Weiher) Date: Thu Nov 3 14:41:50 2005 Subject: Thread Local Storage? Message-ID: > Is there a really fast way to do thread-local storage under Mach? > The only way I can think of would be to use thread_self() > with a hash table that points to the local data you need. > Unfortunately I think this could be far too slow > for my eventual needs, since I *think* thread_self() > is a system call. Just to verify some assumptions: marcel@spock[Tests]cat thread_self_speed.c; cc -O -Wall -o thread_self_speed thread_self_speed.c ; time ./thread_self_speed #import #import int main( int argc, char *argv[] ) { int i,j=0; for (i=0;i<10000000;i++) // that's 10 million { cthread_t thread=cthread_self(); j+=(int)thread; // just to make sure the compiler doesn't } // strength-reduce the call return j; // and doesn't remove it } real 0m6.861s user 0m6.830s sys 0m0.010s <-- virtually no system time marcel@spock[Tests] So using cthreads it looks like getting your own thread is fairly fast ( ~700 nanoseconds on a 233 MHz G3) and doesn't involve the system, which makes sense considering cthreads are a user level threading package. Marcel From kcd at jumpgate.com Tue Feb 16 02:09:01 1999 From: kcd at jumpgate.com (Kenneth C. Dyke) Date: Thu Nov 3 14:41:50 2005 Subject: Thread Local Storage? In-Reply-To: References: Message-ID: <199902161009.CAA04133@babylon5.jumpgate.com> Marcel Weiher wrote: > > Is there a really fast way to do thread-local storage under Mach? > > The only way I can think of would be to use thread_self() > > with a hash table that points to the local data you need. > > Unfortunately I think this could be far too slow > > for my eventual needs, since I *think* thread_self() > > is a system call. > > Just to verify some assumptions: > > marcel@spock[Tests]cat thread_self_speed.c; cc -O -Wall -o > thread_self_speed thread_self_speed.c ; time ./thread_self_speed > > #import > #import > > int main( int argc, char *argv[] ) > { > int i,j=0; > for (i=0;i<10000000;i++) // that's 10 million > { > cthread_t thread=cthread_self(); > j+=(int)thread; // just to make sure the compiler doesn't > } // strength-reduce the call > return j; // and doesn't remove it > } > > real 0m6.861s > user 0m6.830s > sys 0m0.010s <-- virtually no system time > marcel@spock[Tests] > > So using cthreads it looks like getting your own thread is fairly > fast ( ~700 nanoseconds on a 233 MHz G3) and doesn't involve the > system, which makes sense considering cthreads are a user level > threading package. Well, the only problem is that I can't really use cthread_self() because the caller may not be using cthreads, so I'd have to use the real thread_self() routine. Using thread_self() instead of cthread_self() yields: babylon5% time ./threads 52.166s real 4.210s user 47.920s system 99% ./threads Which is more of what I'd expect. Now of course, it'd be interesting to find out exactly what cthread_self() is doing internally to look up it's own cthread_t. -Ken From nberch at db.lv Tue Feb 16 02:36:26 1999 From: nberch at db.lv (Nils Berzins) Date: Thu Nov 3 14:41:50 2005 Subject: Thread Local Storage? Message-ID: <199902161037.MAA05978@saulite> Hi all, A question about BSD sockets in MOSXS... How do you determine the amount of data available for reading on BSD socket. I have not written any networking code with BSD sockets (just OT on MacOS), so my question is: Is there something like: > OTResult OTCountDataBytes(EndpointRef ref, size_t* countPtr); in OpenTransport ? After some digging at 3.00 last night, I found that I should probably use select(). Unfortunately man page info on select() is rather... obscure. It would be very nice, if somebody could point me to a good sample code that uses select() and recv(). Best Regards Nils Berzins From bbum at codefab.com Tue Feb 16 07:42:57 1999 From: bbum at codefab.com (Bill Bumgarner) Date: Thu Nov 3 14:41:50 2005 Subject: Thread Local Storage? References: <199902161037.MAA05978@saulite> Message-ID: <36C99201.E57BD9A3@mail.codefab.com> #ifdef WIN32 #import #elif defined(__MACH__) #import #elif defined(__svr4__) #import #import #endif - (unsigned int) _availableByteCountNonBlocking { #ifdef WIN32 HANDLE nativeHandle = [self nativeHandle]; DWORD lpTotalBytesAvail; BOOL peekSuccess; peekSuccess = PeekNamedPipe(nativeHandle, NULL, 0L, NULL, &lpTotalBytesAvail, NULL); if (peekSuccess == NO) [NSException raise: NSFileHandleOperationException format: @"PeekNamedPipe() NT Err # %d", GetLastError()]; return lpTotalBytesAvail; #elif defined(__MACH__) || defined(__svr4__) int numBytes; int fd = [self fileDescriptor]; if(ioctl(fd, FIONREAD, (char *) &numBytes) == -1) [NSException raise: NSFileHandleOperationException format: @"ioctl() Err # %d", errno]; return numBytes; #else #warning Non-blocking I/O not supported on this platform.... abort(); return nil; #endif } This will work on MacOSX and WIN32 -- likely on Solaris too, though it'll currently #error out.... I added this as a category to NSFileHandle along with: - (NSData *)availableDataNonBlocking; /*" * Returns an NSData object containing all of the currently available data. Does not block if there is no data; returns nil * instead. "*/ { return [self readDataOfLengthNonBlocking: UINT_MAX]; } - (NSData *)readDataToEndOfFileNonBlocking; /*" * Returns an NSData object containing all of the currently available data. Does not block if there is no data; returns nil * instead. Cover for #{-availableDataNonBlocking}. "*/ { return [self readDataOfLengthNonBlocking: UINT_MAX]; } - (NSData *)readDataOfLengthNonBlocking:(unsigned int)length; /*" * Reads up to length bytes of data from the file handle. If no data is available, returns nil. Does not block. "*/ { unsigned int readLength; readLength = [self _availableByteCountNonBlocking]; readLength = (readLength < length) ? readLength : length; if (readLength == 0) return nil; return [self readDataOfLength: readLength]; } b.bum Nils Berzins wrote: > Hi all, > > A question about BSD sockets in MOSXS... > > How do you determine the amount of data available for reading on > BSD socket. I have not written any networking code with BSD sockets > (just OT on MacOS), so my question is: > Is there something like: > > OTResult OTCountDataBytes(EndpointRef ref, size_t* countPtr); > in OpenTransport ? After some digging at 3.00 last night, I found > that I should probably use select(). Unfortunately man page info on > select() is rather... obscure. > It would be very nice, if somebody could point me to a good sample > code that uses select() and recv(). > > Best Regards > Nils Berzins From asen at dti.net Tue Feb 16 09:04:06 1999 From: asen at dti.net (Anders Sensenig) Date: Thu Nov 3 14:41:50 2005 Subject: NSTableView and custom NSCell drawing Message-ID: <9902161204.AA064628@ejungle> Hello! I have an NSTableView connected to an EODataSource. I know how to modify the contents of an NSCell based on the value of its associated EO key in the record -- I've subclassed NSTextFieldCell. Now I want to be able to modify the contents of the NSCell -- based on the value of an EO key that is NOT associated with any column. How do I do that? Thanks. -- Anders From asen at dti.net Tue Feb 16 11:46:17 1999 From: asen at dti.net (Anders Sensenig) Date: Thu Nov 3 14:41:50 2005 Subject: NSTableView and custom NSCell drawing In-Reply-To: <9902161204.AA064628@ejungle> Message-ID: <9902161446.AA178673@ejungle> FWIW solved this... I wrote: > I have an NSTableView connected to an EODataSource. I know how to modify the contents of an NSCell based on the value of its associated EO key in the record -- I've subclassed NSTextFieldCell. Now I want to be able to modify the contents of the NSCell -- based on the value of an EO key that is NOT associated with any column. How do I do that? Thanks. > ... set up a new NSColumn in code, established an EOColumnAssociation to the EODisplayGroup, but did NOT add it to the NSTableView, so it remains invisible. Then got the value for the row/col by sending tableView:objectValueForTableColumn:row: to the NSTableView's dataSource from the delegate's tableView:willUpdateCell:forTableColumn:row method, with the new NSColumn as parameter. -- Anders From bungi at omnigroup.com Tue Feb 16 11:45:35 1999 From: bungi at omnigroup.com (Timothy J. Wood) Date: Thu Nov 3 14:41:50 2005 Subject: Thread Local Storage? Message-ID: <199902161945.LAA12393@mungi.omnigroup.com> A non-text attachment was scrubbed... Name: not available Type: text/enriched Size: 5568 bytes Desc: not available Url : /mailman/archive/macosx-dev/attachments/19990216/423dceda/attachment.bin From kgb at graduate.net Tue Feb 16 12:03:20 1999 From: kgb at graduate.net (KGB Software) Date: Thu Nov 3 14:41:50 2005 Subject: ComboBoxes in a TableColumn Message-ID: Hi, I have a table, and I'd like to make one of the columns have ComboBoxes at each row, rather than just textFields. How do I go about setting this up? And has anyone tried to build an EOAssociation to connect a tableColumn of comboBoxes to a database? Thanks, --John Kuszewski From fabienr at ncmi.com Tue Feb 16 13:09:39 1999 From: fabienr at ncmi.com (Fabien Roy) Date: Thu Nov 3 14:41:50 2005 Subject: Anti spam scheme In-Reply-To: <36C9B993.9675E8A7@inria.fr> References: <199902161430.PAA04932@nez-perce.inria.fr> <36C9B993.9675E8A7@inria.fr> Message-ID: <199902162110.NAA18718@ignem.omnigroup.com> Hi, It's a little bit off topic, but by using the the possibility of SMTP the mail distribution program at omnigroup could reject non verified e-mail addresses. This would increase slightly the traffic or this could be done for any non subscriber sending mail to that list. Here is an example of what I got from the last post of a (potential) spammer from the Omnigroup mailing list. >>> VRFY hotbusiness@olivetti.za <<< 252 Cannot verify user This one is related to MSExchange not implementing the RFCs :-( >>> EXPN hotbusiness@olivetti.za <<< 502 command not implemented Later. Fabien --- Fabien L Roy NationsBanc Montgomery Securities, LLC voice: (704)386-75-76 100 North Tryon Street fax: (704)388-95-64 NC1-007-09-08 Charlotte NC 28255 Beeper: 143-9722 (1-800-946-46-46) or http://www.MobileComm.com/message/ email: fabienr@ncmi.com (NeXT/Mime) Pager-email 1439722@mobilecomm.net From mtrent at msn.fullfeed.com Tue Feb 16 15:27:01 1999 From: mtrent at msn.fullfeed.com (mtrent@msn.fullfeed.com) Date: Thu Nov 3 14:41:50 2005 Subject: ComboBoxes in a TableColumn In-Reply-To: from "KGB Software" at Feb 16, 99 12:28:02 pm Message-ID: <199902162327.RAA09532@shell1.msn.fullfeed.com> "In a previous message you (KGB Software) write:" > > I have a table, and I'd like to make one of the columns have > ComboBoxes at each row, rather than just textFields. How do I > go about setting this up? I've been doing this a lot lately. I find the easiest way is just to allocate a combo box and hook it's cell up as my column's data cell. Here's some sample code; put it whereever you initialize your tableView: _box = [[NSComboBox alloc] init]; [[_box cell] setBordered:NO]; [_box setDataSource:[AttributeComboBoxDataSource sharedInstance]]; [_box setNumberOfVisibleItems:10]; [[tableView tableColumnWithIdentifier:identifier] setDataCell:[_box cell]]; _box is an ivar of type NSComboBox*. identifier is the identifier of the column you're interested in. By removing the _box's border, the control will draw a little cleaner in the column. I'm releasing _box in my object's dealloc method; I haven't looked into releasing it before then. AttributeComboBoxDataSource is just my data source. Your textView datasource treats this column like other text columns: you set and return NSString objects. Enjoy, Mike Trent From brianw at sounds.WA.COM Wed Feb 17 10:34:07 1999 From: brianw at sounds.WA.COM (Brian Willoughby) Date: Thu Nov 3 14:41:50 2005 Subject: Simple Example of pre vs post GUI application In-Reply-To: References: Message-ID: <199902171834.KAA23129@sounds.WA.COM> Andy, [ Does anyone have a favorite example of an application that was [ commnad line based that was converted to a GUI one? If not can [ anyone whip up a quick 5 liner that shows pre GUI code and post [ GUI code? [ I can use Interface Builder like a snap and Instantiate and add [ file to project and all that. Somehow I'm missing that little [ simple something when it comes to this type of conversion. I've been looking for an old magazine article ever since your question came through. Way back when I had just purchased my first NeXTdimension, I found the following article quite helpful in getting me started: OBJECT magazine, 1(4), JAN/FEB 1992, "NeXTstep and me" pp58-66 The article describes creating a NeXTstep program which is a GUI wrapper around the command-line calendar program "cal." This isn't quite what you were asking for, since the author was not working with the cal source code. Instead, he ran the cal program in the background and piped its output into the GUI. This certainly isn't the most efficient way of doing things, especially if you have the source to a command line program (pipes between processes are much slower than messages between ObjC objects), but I think that you might get some good ideas. It certainly started me off in the right direction. Brian Willoughby Software Design: NEXTSTEP, OpenStep, Rhapsody, MacOS X Sound Consulting Apple Enterprise Alliance Partner NeXTmail welcome Apple is the registered trademark of Apple Computer, Inc. and Apple Records From jpb at bofh.creat.cas.ucf.edu Wed Feb 17 09:26:32 1999 From: jpb at bofh.creat.cas.ucf.edu (Joe Block) Date: Thu Nov 3 14:41:50 2005 Subject: Anti spam scheme References: <199902162110.NAA18718@ignem.omnigroup.com> Message-ID: In article <199902162110.NAA18718@ignem.omnigroup.com>, Fabien Roy wrote: > It's a little bit off topic, but by using the the possibility of > SMTP the mail distribution program at omnigroup could reject non > verified e-mail addresses. > This would increase slightly the traffic or this could be done for > any non subscriber sending mail to that list. A lot of sites don't allow vrfy because it lets spammers determine whether a given address is valid. How about restricting posting to list members? If someone wants to post, making them do it from their subscribed address doesn't seem like too much of a hassle to the user. From rcfa at cubiculum.com Wed Feb 17 11:26:46 1999 From: rcfa at cubiculum.com (Ronald C.F. Antony) Date: Thu Nov 3 14:41:50 2005 Subject: Anti spam scheme In-Reply-To: References: Message-ID: <9902171926.AA23919@kannix.cubiculum.com> > How about restricting posting to list members? If someone wants to post, > making them do it from their subscribed address doesn't seem like too much > of a hassle to the user. That's fine if the list software isn't braindead. e.g. there are a few lists that refuse postings from me beause my e-mail goes through UUCP on it's way out of my net, and thus has From uucp even though it has From: rcfa... etc. To get anything listed on these lists I have to rlogin into the gateway computer etc. Real bitch. Given that the list software used by Omni so far can't even honor the Reply-to: field and instead uses the From: field doesn't make me too confident that the current software would do the proper thing... 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 fabienr at ncmi.com Wed Feb 17 11:53:24 1999 From: fabienr at ncmi.com (Fabien Roy) Date: Thu Nov 3 14:41:50 2005 Subject: Anti spam scheme In-Reply-To: References: Message-ID: <199902171954.LAA06572@ignem.omnigroup.com> You wrote: > A lot of sites don't allow vrfy because it lets spammers determine > whether a given address is valid. I did not think of that possibility. :-( > > How about restricting posting to list members? If someone wants to > post, making them do it from their subscribed address doesn't seem > like too much of a hassle to the user. It seems now impossible to protect us from spamming :-( Later. Fabien --- Fabien L Roy NationsBanc Montgomery Securities, LLC voice: (704)386-75-76 100 North Tryon Street fax: (704)388-95-64 NC1-007-09-08 Charlotte NC 28255 Beeper: 143-9722 (1-800-946-46-46) or Hi all, I just ported an old NeXT app to OS X which uses objC++. After converting the sources to conform to the new "extern" behavior I was abe to compile everything just fine, but I had a very strange link error. All Function symbols (NSBeep etc) which came from ObjC frameworks have not been known to the C++ world. So every function produced a link error. This, however, worked in the old days. My current hack uses a "bridge" class where all functions are wrapped in ObjC class methods. This works just fine. So resolving symbols from the ObjC runtime is no problem. Any ideas ? aloha Tomi P.s: I can send the sources ... its a game app ... to whoever wants to take a close look. From cb at df.lth.se Thu Feb 18 01:48:52 1999 From: cb at df.lth.se (Christian Brunschen) Date: Thu Nov 3 14:41:50 2005 Subject: ObjC C++ and the function problem In-Reply-To: <9902180935.AA00635@koala.objectfarm.org> Message-ID: On Thu, 18 Feb 1999, Tomi Engel wrote: > Hi all, > > I just ported an old NeXT app to OS X which uses objC++. > After converting the sources to conform to the new "extern" behavior > I was abe to compile everything just fine, but I had a very strange > link error. > > All Function symbols (NSBeep etc) which came from ObjC frameworks > have not been known to the C++ world. So every function produced a > link error. This, however, worked in the old days. > > My current hack uses a "bridge" class where all functions are > wrapped in ObjC class methods. This works just fine. > So resolving symbols from the ObjC runtime is no problem. > > Any ideas ? are the declarations of these function wrapped in "extern C" ? otherwise, do something like extern "C" { #import } > > aloha > Tomi > > P.s: I can send the sources ... its a game app ... to whoever wants > to take a close look. > // Christian Brunschen From ahoesch at smartsoft.de Thu Feb 18 10:30:41 1999 From: ahoesch at smartsoft.de (Andreas Hoeschler) Date: Thu Nov 3 14:41:50 2005 Subject: OPENSTE 4.2 Mach --> WebObjects 3.5 Differences Message-ID: <9902181830.AA08973@smartsoft.de> A non-text attachment was scrubbed... Name: not available Type: text/enriched Size: 933 bytes Desc: not available Url : /mailman/archive/macosx-dev/attachments/19990218/ad469b00/attachment.bin From ahoesch at smartsoft.de Thu Feb 18 12:13:32 1999 From: ahoesch at smartsoft.de (Andreas Hoeschler) Date: Thu Nov 3 14:41:50 2005 Subject: OPENSTE 4.2 Mach --> WebObjects 3.5 Differences In-Reply-To: <9902181909.AA02919@vviuh221.vvi.com> References: <9902181830.AA08973@smartsoft.de> <9902181909.AA02919@vviuh221.vvi.com> Message-ID: <9902182013.AA10386@smartsoft.de> A non-text attachment was scrubbed... Name: not available Type: text/enriched Size: 1270 bytes Desc: not available Url : /mailman/archive/macosx-dev/attachments/19990218/c23ffb06/attachment.bin -------------- next part -------------- A non-text attachment was scrubbed... Name: console.log Type: application/octet-stream Size: 868 bytes Desc: not available Url : /mailman/archive/macosx-dev/attachments/19990218/c23ffb06/console.obj -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: text/enriched Size: 41 bytes Desc: not available Url : /mailman/archive/macosx-dev/attachments/19990218/c23ffb06/attachment-0001.bin From janos.lobb at yale.edu Thu Feb 18 12:12:26 1999 From: janos.lobb at yale.edu (=?iso-8859-1?Q?J=E1nos?= =?iso-8859-1?Q?_?= =?iso-8859-1?Q?L=F6bb?=) Date: Thu Nov 3 14:41:50 2005 Subject: Newbie question Message-ID: I am working on a project using CG++ from Belmont. /I am new to that too./ It is RAD tool and can produce C++ code for compilation. I would like to know if I want to port to YB or Rhapsody, then should I: 1., Do parallel development of everything in/with YB or 2., Do everything in CG++, create the the c++ code and compile that with the GNU compiler. J?nos J?nos L?bb Tel: 203-737-5204 Yale University Pathology Fax: 203-785-7303 310 Cedar St. Room BML104A janos.lobb@yale.edu New Haven CT 06510 Never take a candid-cookie from a stranger. From ahorovit at apple.com Thu Feb 18 17:00:46 1999 From: ahorovit at apple.com (Alex Horovitz) Date: Thu Nov 3 14:41:50 2005 Subject: Java Developers Journal Readers Choice Awards Message-ID: <199902190057.QAA45496@scv4.apple.com> OK folks, (A) This is NOT an official announcement from Apple in any way shape or form. ---- (B) WebObjects was the first application server on the market, and with thousands of customers, is one of the leading players in the market.Let your voice be heard and vote for the first and the best application server from Apple! Based on their rule number 5: 5) YOU ARE ALLOWED TO VOTE FOR YOUR OWN PRODUCT. PARTICIPATING PRODUCT VENDORS ARE ALLOWED TO ENCOURAGE THEIR USERS TO VOTE FOR THEIR PRODUCTS WITHIN THE RULES AS DESCRIBED ABOVE. So, help WebObjects out here and vote at: http://www.syscon.com/java/readerschoice/nominationform.html The winning products will be announced on SYS-CON's Java Radio during the 1999 JavaOne Conference in San Francisco in June. Lets work together and win one for WebObjects! ---- (C) Please refer to (A) :-) ---- Alex (A random guy who works at Apple on Enterprise products and loves WebObjects) From pekeler at codefab.com Thu Feb 18 17:22:24 1999 From: pekeler at codefab.com (Christian Pekeler) Date: Thu Nov 3 14:41:50 2005 Subject: Java Developers Journal Readers Choice Awards In-Reply-To: <199902190057.QAA45496@scv4.apple.com> Message-ID: <9902182022.AA245217@bass> > http://www.syscon.com/java/readerschoice/nominationform.html Try this http://www.JavaDevelopersJournal.com/java/readerschoice/nominationform.html -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: text/enriched Size: 265 bytes Desc: not available Url : /mailman/archive/macosx-dev/attachments/19990218/7641932b/attachment.bin From tomi at objectfarm.org Fri Feb 19 03:40:55 1999 From: tomi at objectfarm.org (Tomi Engel) Date: Thu Nov 3 14:41:50 2005 Subject: ObjC C++ and the function problem ... solution Message-ID: <9902191137.AA00286@koala.objectfarm.org> Hi, David Young reminded me to check all Appkit and Foundation includes and there it was: One of the AppKit includes was not wrapped inside an extern "Objective-C" { ... } statement. During some changes which I have made on the OS X port it seems like the order of include statements on the C++ side have changed and the "offending" header now was the first to import AppKit. .. and .. bang. Thanx forpushing me into the right direction ... you will be rewarded with a ported "Shanghai" soon. Aloha Tomi From dyoung at vviuh221.vvi.com Fri Feb 19 04:53:48 1999 From: dyoung at vviuh221.vvi.com (David Young) Date: Thu Nov 3 14:41:50 2005 Subject: ObjC C++ and the function problem ... solution References: <9902191137.AA00286@koala.objectfarm.org> Message-ID: <9902191253.AA03816@vviuh221.vvi.com> I always make that mistake too. About every 8.5 months for the last 5 years I e-mail NeXT/Apple to put this: #ifdef __cplusplus extern "Objective-C" { #endif #ifdef __cplusplus } #endif in their header files like every good citizen should do. That would solve the issue. But I think they want to keep their header files pretty looking. Thanks A Bunch! David Young; VVI-DCS http://www.vvi.com From uli at haase-segel.de Fri Feb 19 06:15:57 1999 From: uli at haase-segel.de (Ulrich =?iso-8859-1?Q?K=F6ster?=) Date: Thu Nov 3 14:41:50 2005 Subject: uml and objective c Message-ID: <36CD721C.E554546A@haase-segel.de> *This message was transferred with a trial version of CommuniGate(tm) Pro* any hints, any tools?? thanks in advance Ulrich From pierce at alumni.caltech.edu Fri Feb 19 08:54:12 1999 From: pierce at alumni.caltech.edu (Pierce T. Wetter III) Date: Thu Nov 3 14:41:50 2005 Subject: uml and objective c In-Reply-To: <36CD721C.E554546A@haase-segel.de> Message-ID: At 8:07 AM -0800 2/19/99, Ulrich K?ster wrote: > *This message was transferred with a trial version of CommuniGate(tm) Pro* > any hints, any tools?? > > thanks in advance > > Ulrich If you have a Macintosh, ObjectPlant is pretty cool for UML designs. 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 pierce at twinforces.com Fri Feb 19 09:28:21 1999 From: pierce at twinforces.com (Pierce T. Wetter III) Date: Thu Nov 3 14:41:50 2005 Subject: Help with BSD ports In-Reply-To: <36CD721C.E554546A@haase-segel.de> Message-ID: I'm trying to compile all the Free/Open/NetBSD stuff for Rhapsody. When I'm done, I'll create installer packages for MacOS X, and one of the Linux vendors is going to publish a CD-ROM of everything. Which would be good, considering that there seems to be a lot missing (probably for marketing reasons.) Anyways, I've run into problems, and I was wondering if anyone could help me. While MacOSX is supposed to be BSD compatible, it seems to have a different library structure, and its missing a few things from the Free/Open/NetBSD. In fact, its really BSD 4.4 Lite, i.e. BSD circa 1995. That's both good news and bad news. Its good news because that means that if I do all this work to get everything ported, it will have been worth it. The bad news, is that there doesn't seem to be much documentation on the BSD subsystem. Has anyone ported any of the BSD tools already and can give me a hand? My test case for now is the ftp tool. The current ftp tool can't handle URLs even though FreeBSD has had this since 1997 (though the man page says it can, go figure). Things I've tried: 1. Tweaking the BSD makefiles. This has gotten me the farthest since they have a zillion options controlled by environment variables. I run into problems I can't solve because the library structure is different. For instance, the ftp make complains that there is no libcrt0. Since none of the make files even mention libcrt0 (other then specifying its path), I'm at a loss to figure out how to fix this. I suspect it has something to do with knowledge imbedded in make about what files are needed (which knowledge just happens to be wrong. 2. I've also tried using ProjectBuilder, but it seems more annoying then its worth. First off, it makes about a zillion folders which seems overkill for something with only 8 source files, and then it doesn't work anyways. (I get a whole load of "duplicate definition" errors that are completely incomprehensible. ) My questions: 1. Has anyone done a port, and has helpful suggestions? 2. Has anyone figured out what libraries are what (or where they're documented) in /usr/lib? I could probably fix some builds by just putting in symbolic links to the slightly different named library. 3. Anyone know where or what causes the various libraries to be included? Its it a config file to makedep or something? 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 uli at haase-segel.de Fri Feb 19 09:35:40 1999 From: uli at haase-segel.de (Ulrich =?iso-8859-1?Q?K=F6ster?=) Date: Thu Nov 3 14:41:50 2005 Subject: uml and objective c References: Message-ID: <36CDA0EC.92137218@haase-segel.de> *This message was transferred with a trial version of CommuniGate(tm) Pro* hi pierce ObjectPlant is okay for UML but lacks support for ObjectiveC. Is there no solution for Objective C. Only paper and pencil??? Thanks Ulrich From pierce at alumni.caltech.edu Fri Feb 19 09:50:31 1999 From: pierce at alumni.caltech.edu (Pierce T. Wetter III) Date: Thu Nov 3 14:41:50 2005 Subject: uml and objective c In-Reply-To: <36CDA0EC.92137218@haase-segel.de> References: Message-ID: At 6:35 PM +0100 2/19/99, Ulrich K?ster wrote: > *This message was transferred with a trial version of CommuniGate(tm) Pro* > hi pierce > > ObjectPlant is okay for UML but lacks support for ObjectiveC. You can make your own output templates for ObjectPlant that would generate Objective-C, then it would support it. If you send the author the template, he'll put it in the next revisions. It would probably take you less then a day to do, its pretty simple. If you were hoping for something that would parse Objective-C for reverse engineering, I don't see that happening, though you could write a program that would interrogate all your classes using whatever the Objective-C equivalent of the Java reflection API is. > > Is there no solution for Objective C. Only paper and pencil??? Personally, I find that ALL the UML tools suck! I use paper and pencil until I've got the design mostly final. Then I use a tool to make it pretty. Using a tool encourages you to make minor changes, while using paper and pencil is more conducive to dramatic improvements in the project architecture. Also, I more or less assume that any time I have a one to many association, that I'll have some sort of "collection" that implements the association (i.e. a list, set, whatever), and none of the code generation engines from UML are smart enough to do that as far as I know. Conversely, if I have a one-to-one relationship, its probably a pointer, and I want to set it up in my constructor, and so on. So using the code generation out of any of these UML packages is somewhat of a waste. 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 pekeler at codefab.com Fri Feb 19 13:46:22 1999 From: pekeler at codefab.com (Christian Pekeler) Date: Thu Nov 3 14:41:50 2005 Subject: PBButtons 1.0 Message-ID: <9902191646.AA223131@bass> PBButtons - a bundle for ProjectBuilder. Because I was sick of typing "fb -[NSException raise]" into the launch panel of ProjectBuilder and also don't like the way the build targets have to be selected I created a little bundle which hacks some buttons into the launch and build panel. I think it saves me some time so maybe it's helpful to others, too. The source is here: ftp://ftp.codefab.com/pub/unsupported/PBButtons-1.0-PW-s.tar.gz and a binary compiled for WO4.0/WinNT and Rhapsody/PPC is here: ftp://ftp.codefab.com/pub/unsupported/PBButtons-1.0-PW-b.tar.gz Christian From richs at win.bright.net Fri Feb 19 15:10:03 1999 From: richs at win.bright.net (Rich Schroedel) Date: Thu Nov 3 14:41:50 2005 Subject: Floating menu problem Message-ID: <36CDEF49.7220CF93@win.bright.net> While using Rhapsody DR2, I seem to have accidentally created a floating menu copy of the Workspace Manager menu. It hangs just below the menu bar just a little to the right of where the Workspace Manager menu normally drops down. I suspect it's a floating menu with it's title hid behind the menu bar. It works normally, but I just can't get rid if it. Does anyone have any idea how I can get rid of it short of doing a complete install? Rebooting does no good. -- Rich Schroedel richs@win.bright.net From bbum at codefab.com Fri Feb 19 15:31:22 1999 From: bbum at codefab.com (Bill Bumgarner) Date: Thu Nov 3 14:41:50 2005 Subject: Floating menu problem In-Reply-To: <36CDEF49.7220CF93@win.bright.net> Message-ID: 'defaults remove Workspace ApplicationsMenu' will do the trick. b.bum On Fri, 19 Feb 1999, Rich Schroedel wrote: > > While using Rhapsody DR2, I seem to have accidentally created a floating menu copy of > the Workspace Manager menu. It hangs just below the menu bar just a little to the > right of where the Workspace Manager menu normally drops down. I suspect it's a > floating menu with it's title hid behind the menu bar. It works normally, but I just > can't get rid if it. Does anyone have any idea how I can get rid of it short of doing > a complete install? Rebooting does no good. > > -- > Rich Schroedel > richs@win.bright.net > > > From rcfa at cubiculum.com Fri Feb 19 15:32:16 1999 From: rcfa at cubiculum.com (Ronald C.F. Antony) Date: Thu Nov 3 14:41:51 2005 Subject: Floating menu problem In-Reply-To: <36CDEF49.7220CF93@win.bright.net> References: <36CDEF49.7220CF93@win.bright.net> Message-ID: <9902192332.AA03261@kannix.cubiculum.com> you wrote: > While using Rhapsody DR2, I seem to have accidentally created a floating > menu copy of the Workspace Manager menu. It hangs just below the menu bar > just a little to the right of where the Workspace Manager menu normally > drops down. I suspect it's a floating menu with it's title hid behind the > menu bar. It works normally, but I just can't get rid if it. Does anyone > have any idea how I can get rid of it short of doing a complete install? > Rebooting does no good. The location of this menu is probably saved in the defaults database. You can edit that database with the corresponging tools. There is however one caveat: Workspace possibly updates the defaults as you log out, which means what you changes might be overwritten and thus trashed. If that's the case, you must change the defaults database while not being logged in with the the Workspace. You can do that by using the root account, and then su-ing to your account, or alternatively, rlogin into your account from another machine on the net that doesn't share the defaults. 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 lux at umich.edu Fri Feb 19 15:57:40 1999 From: lux at umich.edu (Joshua Marker) Date: Thu Nov 3 14:41:51 2005 Subject: Floating menu problem In-Reply-To: <9902192332.AA03261@kannix.cubiculum.com> Message-ID: On Fri, 19 Feb 1999, Ronald C.F. Antony wrote: > > While using Rhapsody DR2, I seem to have accidentally created a floating > > menu copy of the Workspace Manager menu. It hangs just below the menu bar > > just a little to the right of where the Workspace Manager menu normally > > drops down. I suspect it's a floating menu with it's title hid behind the > > menu bar. It works normally, but I just can't get rid if it. Does anyone > > have any idea how I can get rid of it short of doing a complete install? > > Rebooting does no good. > > The location of this menu is probably saved in the defaults database. > You can edit that database with the corresponging tools. There is Alternatively, one can tear off a new menu, which will replace the old one. Put it somewhere you can close it, and do so. --- Joshua (lux) * Kalamazoo: 616 / 337.7479 Sorting out my life. . . .in O(n lg n). "Stop it, Ray. You're scaring the straights." --- From sanguish at digifix.com Fri Feb 19 20:58:31 1999 From: sanguish at digifix.com (Scott Anguish) Date: Thu Nov 3 14:41:51 2005 Subject: YB Windows linking problems with bundles In-Reply-To: <199902190720.CAA00450@digifix.com> References: <199902190720.CAA00450@digifix.com> Message-ID: <199902200458.XAA04505@digifix.com> I'm encountering a compiling/linking problem on Yellow Box windows that isn't a problem on DR2.. I've got an application that has three layers User Interface | -------------- basic data classes ------ ------ ------ ------ bundle bundle bundle bundle The bundles are all the individual tools... they use the classes in the basic data classes layers. on Mach I can just find all the bundles, load them, and instantiate the objects as required. No problem.. However, compiling on Windows using the default make settings, this just doesn't work.. I'm getting link errors for the superclass of the principal object in the bundle, as well as any other references to the basic data clases. I'm hoping that there is a quick makefile fix on Windows so I don't have to make the basic data classes into a framework (which complicates installation for my users). Any ideas? From kcd at jumpgate.com Fri Feb 19 23:16:02 1999 From: kcd at jumpgate.com (Kenneth C. Dyke) Date: Thu Nov 3 14:41:51 2005 Subject: YB Windows linking problems with bundles In-Reply-To: <199902200458.XAA04505@digifix.com> References: <199902200458.XAA04505@digifix.com> Message-ID: <199902200716.XAA08974@babylon5.jumpgate.com> A non-text attachment was scrubbed... Name: not available Type: text/enriched Size: 864 bytes Desc: not available Url : /mailman/archive/macosx-dev/attachments/19990219/ba393eb4/attachment.bin From bbum at codefab.com Fri Feb 19 23:21:55 1999 From: bbum at codefab.com (Bill Bumgarner) Date: Thu Nov 3 14:41:51 2005 Subject: YB Windows linking problems with bundles In-Reply-To: <199902200458.XAA04505@digifix.com> Message-ID: All of the classes used by the bundles must be dynamically linkable and must be present in the runtime before you can load a bundle [under NT-- this is not the case on mach]. As such, you have to make sure that all symbols that your bundle uses are defined in frameworks that your bundle links against (or go through the trouble of creating a standalone .dll and link against that). A pain, isn't it. It means that you basically can't have a .app that defines any classes or symbols from which your bundle either inherits (or references via cateogories or, even, via [ClassName alloc]). As long as you don't refer to any symbols (methods are OK because of ObjC's dynamic linker)-- including classes-- directly, you can successfully link and subsequently dyna-load a bundle. However, it becomes a bit tedious to use NSClassFromString() *everywhere*. As well, it means you can't directly use extern NSString *ExceptionOrNotificationName variables. I have basically resorted to creating every app as an Aggregate project (or similar via cfbuild... an entirely multiscriptshell based build thingy that allows one to build frameworks anywhere for any path-- if anyone wants to contribute to making it something other than a quick&dirty hack, i would really like the help) that contains a Framework and an Application project-- the Application links against the framework and is basically a one liner: ...main(...) { ....pool stuff... ....[AppController main]; ....un pool... } And have +main do the same as the normal appkit main(). This way, all bundles on all platforms simply link against the Apple framework. It is a pain to build for debugging purposes until you tweak environmant variables and make variables to do the right thing, but it works on all platforms and lets you not deal with the sometimes goofy NT dyna-linker. b.bum On Fri, 19 Feb 1999, Scott Anguish wrote: > I'm encountering a compiling/linking problem on Yellow Box windows that > isn't a problem on DR2.. > > I've got an application that has three layers > > > User Interface > | > -------------- > basic data classes > ------ ------ ------ ------ > bundle bundle bundle bundle > > > The bundles are all the individual tools... they use the classes in the > basic data classes layers. > > on Mach I can just find all the bundles, load them, and instantiate the > objects as required. No problem.. > > > However, compiling on Windows using the default make settings, this just > doesn't work.. I'm getting link errors for the superclass of the principal > object in the bundle, as well as any other references to the basic data > clases. > > I'm hoping that there is a quick makefile fix on Windows so I don't have to > make the basic data classes into a framework (which complicates installation > for my users). > > Any ideas? > > > > From sanguish at digifix.com Fri Feb 19 23:31:17 1999 From: sanguish at digifix.com (Scott Anguish) Date: Thu Nov 3 14:41:51 2005 Subject: YB Windows linking problems with bundles In-Reply-To: References: Message-ID: <199902200731.CAA05258@digifix.com> Bill Bumgarner wrote: > All of the classes used by the bundles must be dynamically linkable and > must be present in the runtime before you can load a bundle [under NT-- > this is not the case on mach]. > and Kenneth C. Dyke" wrote >Make the basic data classes into a framework. I think that's the >only way for it to work on Windoze. :P Well.. this sucks.. Is there any way on Windows to have your frameworks contained within the app wrapper and still found? Hmmm does this work on Mach (OSXS is my only concern.. not OS) From uli at haase-segel.de Sat Feb 20 00:58:35 1999 From: uli at haase-segel.de (Ulrich =?iso-8859-1?Q?K=F6ster?=) Date: Thu Nov 3 14:41:51 2005 Subject: uml and objective c References: Message-ID: <36CE793C.52FB0914@haase-segel.de> *This message was transferred with a trial version of CommuniGate(tm) Pro* hi pierce Okay , i take a second lok at ObjectPlant. Thanks Ulrich From ahorovit at apple.com Sat Feb 20 05:14:09 1999 From: ahorovit at apple.com (Alex Horovitz) Date: Thu Nov 3 14:41:51 2005 Subject: Readers Choice Awards UPDATE Message-ID: <199902201311.FAA54096@scv4.apple.com> Helping WebObjects win the Java Developers Journal Readers Choice Award for best Application Server: (A) This is NOT an official announcement from Apple in any way shape or form. (B) Thank you to everyone who has voted so far! In 24 hours we have gone from ~15 votes to 306. We are still behind IBM (426) and Silver Stream (422) so we need a continued push from our user base! (C) If you haven't yet had the chance to place your vote for WebObjects, I am asking you again as a personal favor please do so as soon as possible. You can turn to: http://www.javadevelopersjournal.com/java/readerschoice/nominationform.html Remember, we are trying to generate a win for WebObjects which is a big win for this technology and Apple in the enterprise space! (D) If you would like to see the standings, please point your browser to: http://www.javadevelopersjournal.com/java/READERSCHOICE/liveupdateappserver2.cfm (E) see (A) :-) Alex (A random Apple employee who works on WebObjects with the Apple Enterprise Software team) From KWong at gects.ge.com Sat Feb 20 09:39:13 1999 From: KWong at gects.ge.com (Wong, Kai (CAP, ITS, CA)) Date: Thu Nov 3 14:41:51 2005 Subject: YB Windows linking problems with bundles Message-ID: <18D74E90607CD011A8EE0000F830950F02603C77@msx02torctmsge.gects.ge.com> > Scott, > You can set the linker "OTHER_LDFLAGS = -undefined suppress" in > your Makefile.preamble. Make sure you only set the flag for WIN32 > only. Hope that helps! > > > Kai > > Tel:+1 905 507-5341 > eCommerce > GE Capital IT Solutions > > ---------- > From: Scott Anguish[SMTP:sanguish@digifix.com] > Reply To: sanguish@digifix.com > Sent: Saturday, February 20, 1999 12:04 AM > To: Multiple recipients of list > Subject: YB Windows linking problems with bundles > > I'm encountering a compiling/linking problem on Yellow Box > windows that > isn't a problem on DR2.. > > I've got an application that has three layers > > > User Interface > | > -------------- > basic data classes > ------ ------ ------ ------ > bundle bundle bundle bundle > > > The bundles are all the individual tools... they use the classes > in the > basic data classes layers. > > on Mach I can just find all the bundles, load them, and > instantiate the > objects as required. No problem.. > > > However, compiling on Windows using the default make settings, > this just > doesn't work.. I'm getting link errors for the superclass of the > principal > object in the bundle, as well as any other references to the > basic data > clases. > > I'm hoping that there is a quick makefile fix on Windows so I > don't have to > make the basic data classes into a framework (which complicates > installation > for my users). > > Any ideas? > > > > From richs at win.bright.net Sat Feb 20 15:22:04 1999 From: richs at win.bright.net (Rich Schroedel) Date: Thu Nov 3 14:41:51 2005 Subject: Floating menu problem References: <36CDEF49.7220CF93@win.bright.net> Message-ID: <36CF439A.FD5D1288@win.bright.net> Thanks for all the helpful replies. For the record, creating a new floating menu and then closing it does flush the "stuck" menu. -- Rich Schroedel richs@win.bright.net From respass at htl.com Sat Feb 20 16:42:49 1999 From: respass at htl.com (Marc Respass) Date: Thu Nov 3 14:41:51 2005 Subject: IMAP email client for MXS? Message-ID: <199902210043.QAA14189@ignem.omnigroup.com> Hi, With the impending release of Mac OS X Server, I am wondering if there is a nice IMAP client out there. I really need an email program that handles multiple accounts and does POP3 and IMAP4. I was looking at the JavaMail API and that may be what I need. I was thinking of writing an IMAP client if there isn't one. Anyone know? Might be a fun project. Since it will be my first YB project, I'll probably make it free and open source. Thanks --Marc Respass From pekeler at codefab.com Sat Feb 20 16:50:43 1999 From: pekeler at codefab.com (Christian Pekeler) Date: Thu Nov 3 14:41:51 2005 Subject: IMAP email client for MXS? In-Reply-To: <199902210043.QAA14189@ignem.omnigroup.com> Message-ID: <9902201950.AA433404@bass> "Marc Respass" wrote: > With the impending release of Mac OS X Server, I am wondering if there is a > nice IMAP client out there. There is: PopOver -->> http://www.eidolonsoftware.com/ Christian From mtrent at msn.fullfeed.com Sat Feb 20 19:06:23 1999 From: mtrent at msn.fullfeed.com (mtrent@msn.fullfeed.com) Date: Thu Nov 3 14:41:51 2005 Subject: IMAP email client for MXS? In-Reply-To: <199902210043.QAA14189@ignem.omnigroup.com> from "Marc Respass" at Feb 20, 99 04:47:59 pm Message-ID: <199902210306.VAA00400@shell1.msn.fullfeed.com> MailViewer.app will support IMAP and multiple accounts. MDT "In a previous message you (Marc Respass) write:" > > Hi, > > With the impending release of Mac OS X Server, I am wondering if there is a > nice IMAP client out there. I really need an email program that handles > multiple accounts and does POP3 and IMAP4. I was looking at the JavaMail API > and that may be what I need. I was thinking of writing an IMAP client if > there isn't one. Anyone know? Might be a fun project. Since it will be my > first YB project, I'll probably make it free and open source. > > Thanks > > --Marc Respass > > From sanguish at digifix.com Sat Feb 20 20:25:13 1999 From: sanguish at digifix.com (Scott Anguish) Date: Thu Nov 3 14:41:51 2005 Subject: IMAP email client for MXS? In-Reply-To: <199902210306.VAA00400@shell1.msn.fullfeed.com> References: <199902210306.VAA00400@shell1.msn.fullfeed.com> Message-ID: <199902210425.XAA07561@digifix.com> mtrent@msn.fullfeed.com wrote: > MailViewer.app will support IMAP and multiple accounts. This is good news... Now all we need is published bundle API (even it its unsupported)... From dyoung at vviuh221.vvi.com Sun Feb 21 05:52:17 1999 From: dyoung at vviuh221.vvi.com (David Young) Date: Thu Nov 3 14:41:51 2005 Subject: gnu obj-c && DO extensions? References: <199902210425.XAA07561@digifix.com> Message-ID: <9902211352.AA06163@vviuh221.vvi.com> Does anyone know: (a) Does the gnu objective-c compiler and run-time (the one supplied directly from the GNU project, not Apple) support the distributed object (DO) protocol extensions (like in out, bycopy, typedstream encoding extensions, etc.)? (b) if (a) is YES, then where is a starting URL so I can find more info.? (c) If (a) is INDEFINITE, does anyone know of a better mail list to ask this question? Thanks A Bunch! David Young; VVI-DCS http://www.vvi.com From richard at brainstorm.co.uk Sun Feb 21 06:44:31 1999 From: richard at brainstorm.co.uk (Richard Frith-Macdonald) Date: Thu Nov 3 14:41:51 2005 Subject: gnu obj-c && DO extensions? In-Reply-To: <9902211352.AA06163@vviuh221.vvi.com> References: <9902211352.AA06163@vviuh221.vvi.com> Message-ID: <9902211444.AA01430@tiptree.brainstorm.co.uk> On Sun, 21 Feb 1999 05:47:43 -0800 (PST), dyoung@vviuh221.vvi.com wrote: > (a) Does the gnu objective-c compiler and run-time (the one supplied > directly from the GNU project, not Apple) support the distributed object (DO) > protocol extensions (like in out, bycopy, typedstream encoding extensions, > etc.)? DO protocol extensions - 'bycopy', 'in', 'out', 'oneway' - have been supported for years, 'byref' is supported in recent (egcs) versions of the compiler. As for the typed stream stuff - the code is in there and might work - but I don't really know since it is really obsolete and not used by anbody that I know of! > (b) if (a) is YES, then where is a starting URL so I can find more info.? I don't know of a URL - try looking at the compiler source. For information about usage of DO you could do a lot worse than looking at Apples documentation (which mostly applies to GNUstep DO too). From dyoung at vviuh221.vvi.com Sun Feb 21 07:07:30 1999 From: dyoung at vviuh221.vvi.com (David Young) Date: Thu Nov 3 14:41:51 2005 Subject: gnu obj-c && DO extensions? References: <9902211352.AA06163@vviuh221.vvi.com> <9902211444.AA01430@tiptree.brainstorm.co.uk> Message-ID: <9902211507.AA06241@vviuh221.vvi.com> Richard (and others), Ack! I asked only part of the question I need answers for ... here is at least another question to more complete the issue: Does the GNU-Foundation Kit look-alike (from gnustep, non-Apple supplied) support the NSCoder DO-related subclasses (NSPortCoder) so objects are encoded and decoded over DO connections? (hence the questions about the typedstream encoding extensions). Thanks A Bunch! David Young; VVI-DCS http://www.vvi.com From dtm at vramp.net Sun Feb 21 06:59:01 1999 From: dtm at vramp.net (David T. McWherter) Date: Thu Nov 3 14:41:51 2005 Subject: gnu obj-c && DO extensions? In-Reply-To: <9902211444.AA01430@tiptree.brainstorm.co.uk> Message-ID: > On Sun, 21 Feb 1999 05:47:43 -0800 (PST), dyoung@vviuh221.vvi.com wrote: > > (a) Does the gnu objective-c compiler and run-time (the one supplied > > directly from the GNU project, not Apple) support the distributed object (DO) > > protocol extensions (like in out, bycopy, typedstream encoding extensions, > > etc.)? > > DO protocol extensions - 'bycopy', 'in', 'out', 'oneway' - have been > supported for years, 'byref' is supported in recent (egcs) versions of the > compiler. What? I submitted a patch for 2.7.2 that gave the Obj-C GCC module the ability to recognize some of those keywords - barely 1 year ago. Prior to that, the Obj-C frontend would balk on those keywords, let alone make use of them (When I finished, they didn't do anything with the keywords, just accepted them as markers on the appropriate nodes in the parse tree). -David -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- David T. McWherter dtm@waterw.com From richard at brainstorm.co.uk Sun Feb 21 07:31:29 1999 From: richard at brainstorm.co.uk (Richard Frith-Macdonald) Date: Thu Nov 3 14:41:51 2005 Subject: gnu obj-c && DO extensions? In-Reply-To: <9902211507.AA06241@vviuh221.vvi.com> References: <9902211352.AA06163@vviuh221.vvi.com> <9902211444.AA01430@tiptree.brainstorm.co.uk> <9902211507.AA06241@vviuh221.vvi.com> Message-ID: <9902211531.AA01511@tiptree.brainstorm.co.uk> On Sun, 21 Feb 99 10:07:30 -0500, dyoung@vviuh221.vvi.com wrote: > Ack! I asked only part of the question I need answers for ... here is at > least another question to more complete the issue: > > Does the GNU-Foundation Kit look-alike (from gnustep, non-Apple supplied) > support the NSCoder DO-related subclasses (NSPortCoder) so objects are > encoded and decoded over DO connections? (hence the questions about the > typedstream encoding extensions). Yes - but it doesn't use the typedstream extensions of the objc runtime to do it. I didn't think that Apples OPENSTEP/MacOS-X implementation used those extensions either. From richard at brainstorm.co.uk Sun Feb 21 07:45:25 1999 From: richard at brainstorm.co.uk (Richard Frith-Macdonald) Date: Thu Nov 3 14:41:51 2005 Subject: gnu obj-c && DO extensions? In-Reply-To: References: Message-ID: <9902211545.AA01529@tiptree.brainstorm.co.uk> On Sun, 21 Feb 1999 09:59:01 -0500 (EST), dtm@vramp.net wrote: > > On Sun, 21 Feb 1999 05:47:43 -0800 (PST), dyoung@vviuh221.vvi.com wrote: > > > (a) Does the gnu objective-c compiler and run-time (the one supplied > > > directly from the GNU project, not Apple) support the distributed object > > > (DO) protocol extensions (like in out, bycopy, typedstream encoding > > > extensions, etc.)? > > > > DO protocol extensions - 'bycopy', 'in', 'out', 'oneway' - have been > > supported for years, 'byref' is supported in recent (egcs) versions of the > > compiler. > > What? I submitted a patch for 2.7.2 that gave the Obj-C GCC > module the ability to recognize some of those keywords - barely > 1 year ago. Prior to that, the Obj-C frontend would balk > on those keywords, let alone make use of them (When I finished, > they didn't do anything with the keywords, just accepted them as > markers on the appropriate nodes in the parse tree). Well, I guess you were a little late - that would have been a short while after gcc-2.8.0 was released, and a few years after most of the world started using gcc-2.7.2.3 :-) The earliest reference I can find is April 1993 when the compiler was modified so that "`in', `out', `inout', `bycopy', and `oneway' are keywords only in protocol declarations". I don't know how well this stuff worked back then, but the 2.7.2.3 compiler implementation was fine back when I started using GNUstep Distributed Objects in 1996. From robnewberry at grouplogic.com Sun Feb 21 09:44:34 1999 From: robnewberry at grouplogic.com (Rob Newberry) Date: Thu Nov 3 14:41:51 2005 Subject: IMAP email client for MXS? In-Reply-To: <199902210425.XAA07561@digifix.com> Message-ID: >> MailViewer.app will support IMAP and multiple accounts. > > This is good news... This IS good news. But is it TRUE? How do we know this will be the case? Did someone just now violate NDA, or has Apple stated this somewhere? Did they once again just forget about telling developers? Whatever the reason -- I sure wish Mac OS X Server would arrive, so I could see for myself if it had an IMAP client. I'm looking forward to something better than pine (though pine is pretty hard to beat). Rob --------------------------------------------------------------------- Rob Newberry Director of Fajita Technology Group Logic, Inc. From Scott_Harrison at apple.com Sun Feb 21 13:51:57 1999 From: Scott_Harrison at apple.com (Scott Harrison) Date: Thu Nov 3 14:41:51 2005 Subject: IMAP email client for MXS? In-Reply-To: Message-ID: <9902211651.AA573431@mage> Rob Newberry made my mailer see: -> -> >> MailViewer.app will support IMAP and multiple accounts. -> > -> > This is good news... -> -> This IS good news. But is it TRUE? -> -> How do we know this will be the case? Did someone just now violate NDA, -> or has Apple stated this somewhere? Did they once again just forget about -> telling developers? -> -> Whatever the reason -- I sure wish Mac OS X Server would arrive, so I -> could see for myself if it had an IMAP client. I'm looking forward to -> something better than pine (though pine is pretty hard to beat). -> -> Rob -> -> --------------------------------------------------------------------- -> Rob Newberry -> Director of Fajita Technology -> Group Logic, Inc. -> -> One can always use PopOver 2 which is a nice mail application that allows multiple POP/IMAP sources. Admittedly, I am only using 3 IMAP sources at the moment, but it works fine. There is a published API for bundles which allow for the numerous bundles that have already been written for it. It works on all the supported platforms (yes, that means that one can read NeXTmail on Windows), handles MIME, etc. Check out http://www.eidolonsoftware.com and take it for a spin. And no, I do not get paid for using it. Scott Harrison From cmh at greendragon.com Sun Feb 21 14:50:36 1999 From: cmh at greendragon.com (Chris Hanson) Date: Thu Nov 3 14:41:51 2005 Subject: IMAP email client for MXS? In-Reply-To: <9902211651.AA573431@mage> Message-ID: Will there be a system-level email API on Mac OS X, so developers don't continue to reinvent the wheel when it comes to mail transport and access? (Actually, will there be a broad base of system-level Internet APIs? I've heard about an NSURL class, which seems like a good start, but is it all or will there be more?) From kelleysoft at pressenter.com Sun Feb 21 18:12:41 1999 From: kelleysoft at pressenter.com (Greg Betzel) Date: Thu Nov 3 14:41:51 2005 Subject: Sharing ObjC headers In-Reply-To: <9902211651.AA573431@mage> Message-ID: This may be a silly question, but here goes anyway... Is there any way for two ObjC header files to #include each-other? For example, in file sourceA.h you'd have #include "sourceB.h", and in file sourceB.h you'd have #include "sourceA.h". The source files are for a series of objects that communicate with each-other rather closely. Normally this wouldn't be a problem, at the worst you'd use id identifiers & call methods with performSelector:. But in this case I'm doing some fairly complex things & would like the type-checking & assurances of calling an object normally. There are other alternatives, like using notifications or even protocols, but I'd like to keep things as clean & simple as possible. Any guesses? Thanks in advance. :) Greg Betzel From ahorovit at apple.com Sun Feb 21 18:48:15 1999 From: ahorovit at apple.com (Alex Horovitz) Date: Thu Nov 3 14:41:51 2005 Subject: JDJ Reader's Choice Award UPDATE Message-ID: <199902220246.SAA20978@scv3.apple.com> Helping WebObjects win the Java Developers Journal Readers Choice Award for best Application Server: (A) This is NOT an official announcement from Apple in any way shape or form. (B) At 9:47 EST we are tied with Silver Stream for second place and pushing hard towards first! Thank you to everyone who has voted so far! In 48 hours we have gone from ~15 votes to 424. We are still behind IBM (431) and Silver Stream (424) so we need a continued push from our user base! (C) If you haven't yet had the chance to place your vote for WebObjects, I am asking you again as a personal favor please do so as soon as possible. You can turn to: http://www.javadevelopersjournal.com/java/readerschoice/nominationform.html Remember, we are trying to generate a win for WebObjects which is a big win for this technology and Apple in the enterprise space! ENCOURAGE EVERYONE YOU KNOW WHO LOVES WEBOBJECTS TO DO THIS ASAP. I'm sure there are people out there we have yet to reach...b (D) If you would like to see the standings, please point your browser to: http://www.javadevelopersjournal.com/java/READERSCHOICE/liveupdateappserver2.cfm (E) Thank you, thank you, thank you! (F) See (A) -Alex (A VERY HAPPY random Apple employee who is really excited about how the WebObjects/Apple user base has responded to these pleas for support! Anyone have any WebObjects questions they need answered?) From cwolf at wolfware.com Sun Feb 21 19:01:04 1999 From: cwolf at wolfware.com (Christopher Wolf) Date: Thu Nov 3 14:41:51 2005 Subject: Sharing ObjC headers Message-ID: <199902220259.VAA03856@mail.rdu.bellsouth.net> To avoid this sort of circular reference use the @class directive to forward declare the classes rather than #importing the header files. (@class is documented in the online Obj-C manual) In fact I go so far as to suggest (and follow) the following rule for header files: - Only #import the superclass header and any protocol declarations that are needed. Declare all other classes referenced within the header file using the @class directive. In addition to avoiding the possibility of circular references this can also reduce compile time and under NT greatly reduce the size of your debug images. - Chris > From: Greg Betzel > Date: 1999-02-21 18:13:05 -0500 > To: Multiple recipients of list > Subject: Sharing ObjC headers > X-UIDL: 95a0e2551cf174d65339931f391bb9ec > 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/ > > > This may be a silly question, but here goes anyway... > > Is there any way for two ObjC header files to #include each-other? > > For example, in file sourceA.h you'd have #include "sourceB.h", and in > file > sourceB.h you'd have #include "sourceA.h". > > The source files are for a series of objects that communicate with > each-other rather closely. Normally this wouldn't be a problem, at the > worst you'd use id identifiers & call methods with performSelector:. But > in this case I'm doing some fairly complex things & would like the > type-checking & assurances of calling an object normally. There are other > alternatives, like using notifications or even protocols, but I'd like to > keep things as clean & simple as possible. > > Any guesses? Thanks in advance. :) > > Greg Betzel > > From ajetha at tacticalstep.com Sun Feb 21 19:57:11 1999 From: ajetha at tacticalstep.com (Alykhan Jetha) Date: Thu Nov 3 14:41:51 2005 Subject: IMAP email client for MXS? References: Message-ID: <36D0D597.B1A3CFDB@tacticalstep.com> If WO4.0 is any indication, there will be support to send messages at least. Here is the header file to NSMailDelivery in Message.framework. /* NSMailDelivery.h Message Framework Copyright (c) 1997, Apple Computer, Inc. All rights reserved. */ #import @interface NSMailDelivery : NSObject { } + (BOOL)deliverMessage: (NSAttributedString *)messageBody headers: (NSDictionary *)messageHeaders format: (NSString *)messageFormat protocol: (NSString *)deliveryProtocol; // // messageBody is the actual contents of the message. If messageFormat is NSASCIIMailFormat, // any rich text or attachments in messageBody will be lost // // messageHeaders contains all of the values that will appear in the message header, // for example, "To", "From", "Subject", etc. If the value is an array, the elements // in the array will converted to a comma seperated list of values in the message header // // messageFormat can be either NSASCIIMailFormat or NSMIMEMailFormat // // deliveryProtocol can be either NSSendmailDeliveryProtocol or NSSMTPDeliveryProtocol // or nil to use the platform's default deliveryProtocol (this is the preferred choice). // On Mach, the default is Sendmail and on Windows the default is SMTP. // + (BOOL)deliverMessage:(NSString *)messageBody subject:(NSString *)messageSubject to:(NSString *)destinationAddress; // // messageBody is the plain text contents of the message // // messageSubject is the subject of the message // // destinationAddress is the recipient's email address, e.g. johndoe@apple.com // @end // This is to control declarations of exported variables and // functions on Windows. #if defined(WIN32) #undef MESSAGE_EXPORT #if defined(NSBUILDINGMESSAGE) #define MESSAGE_EXPORT __declspec(dllexport) extern #else #define MESSAGE_EXPORT __declspec(dllimport) extern #endif #if !defined(MESSAGE_IMPORT) #define MESSAGE_IMPORT __declspec(dllimport) extern #endif #endif #if !defined(MESSAGE_EXPORT) #define MESSAGE_EXPORT extern #endif #if !defined(MESSAGE_IMPORT) #define MESSAGE_IMPORT extern #endif MESSAGE_EXPORT NSString *NSMIMEMailFormat; MESSAGE_EXPORT NSString *NSASCIIMailFormat; // String constants used to specify the format in the // +deliverMessage:headers:format:protocol: method. MESSAGE_EXPORT NSString *NSSendmailDeliveryProtocol; MESSAGE_EXPORT NSString *NSSMTPDeliveryProtocol; // String constants used to specify the delivery protocol // in the +deliverMessage:headers:format:protocol: method. Chris Hanson wrote: > Will there be a system-level email API on Mac OS X, so developers don't > continue to reinvent the wheel when it comes to mail transport and access? > > (Actually, will there be a broad base of system-level Internet APIs? I've > heard about an NSURL class, which seems like a good start, but is it all or > will there be more?) -- Alykhan Jetha (AJ) TacticalStep Inc., Email: mailto:ajetha@tacticalstep.com Phone: 905-831-2643 ext 1 Fax: 905-837-2546 From ajetha at tacticalstep.com Sun Feb 21 20:03:15 1999 From: ajetha at tacticalstep.com (Alykhan Jetha) Date: Thu Nov 3 14:41:51 2005 Subject: JDJ Reader's Choice Award UPDATE References: <199902220246.SAA20978@scv3.apple.com> Message-ID: <36D0D703.6CF0D6DB@tacticalstep.com> As of 10:59pm EST we are tied for 1st place with IBM at 431 votes. I say we need about 20 more votes to solidify 1st. I've been calling people that are not on these lists to make sure they vote! I'm sure each of you can think of a couple of people to call ... Alex Horovitz wrote: > Helping WebObjects win the Java Developers Journal Readers Choice Award for > best Application Server: > > (A) This is NOT an official announcement from Apple in any way shape or form. > > (B) At 9:47 EST we are tied with Silver Stream for second place and pushing > hard towards first! Thank you to everyone who has voted so far! In 48 hours we > have gone from ~15 votes to 424. We are still behind IBM (431) and Silver > Stream (424) so we need a continued push from our user base! > > (C) If you haven't yet had the chance to place your vote for WebObjects, I am > asking you again as a personal favor please do so as soon as possible. You can > turn to: > > http://www.javadevelopersjournal.com/java/readerschoice/nominationform.html > > Remember, we are trying to generate a win for WebObjects which is a big win > for this technology and Apple in the enterprise space! > > ENCOURAGE EVERYONE YOU KNOW WHO LOVES WEBOBJECTS TO DO THIS ASAP. I'm sure > there are people out there we have yet to reach...b > > (D) If you would like to see the standings, please point your browser to: > > http://www.javadevelopersjournal.com/java/READERSCHOICE/liveupdateappserver2.cfm > > (E) Thank you, thank you, thank you! > > (F) See (A) > > -Alex > > (A VERY HAPPY random Apple employee who is really excited about how the > WebObjects/Apple user base has responded to these pleas for support! Anyone > have any WebObjects questions they need answered?) -- Alykhan Jetha (AJ) TacticalStep Inc., Email: mailto:ajetha@tacticalstep.com Phone: 905-831-2643 ext 1 Fax: 905-837-2546 From jman at clubi.ie Mon Feb 22 05:56:22 1999 From: jman at clubi.ie (justin armstrong) Date: Thu Nov 3 14:41:51 2005 Subject: ObjC/Java bridge problems References: <36CF439A.FD5D1288@win.bright.net> Message-ID: <36D16206.DAA754AB@clubi.ie> Hi, I'm having a lot of problems trying to wrap some objective-c code so it can be called from java. The documentation makes it sound a lot easier than it is :-( I'll describe what I'm doing and maybe someone can point out where I'm going wrong or else just tell me a completely different way of doing it. I first create a framework project that contains the code I want to wrap. When I compile this it produces a framework bundle. Following the documentation, I then create a wrapper project and add the header from my framework (there's only one class in my framework), and the framework bundle itself (oddly enough, it seems to compile fine without it). I'm reasonably sure my .jobs file is okay as well. When this is compiled it produces a .dylib and some stuff in the .build directory. I'm assuming that the only thing relevant to me in this directory is the derived java source file which contains the native stubs (i don't know if this is the right terminology). I add this file and the framework bundle to the java application project that i want to call the native code. This compiles fine, but when it runs it complains about either not being able to find the framework or the dylib. So where do i put all this stuff? I assume that the framework bundle goes in ~/Library/Frameworks/ but where does the dylib go that the wrapper project produced? It would seem like a good place for it would be inside the framework bundle, but moving it inside there doesn't seem to help. Any ideas? Cheers, Justin Armstrong From jman at clubi.ie Mon Feb 22 10:56:28 1999 From: jman at clubi.ie (justin armstrong) Date: Thu Nov 3 14:41:51 2005 Subject: ObjC/Java bridge problems References: <36D16206.DAA754AB@clubi.ie> Message-ID: <36D1A85C.ADE43D8D@clubi.ie> justin armstrong wrote: > Hi, > I'm having a lot of problems trying to wrap some objective-c code so it can be called > from java. > The documentation makes it sound a lot easier than it is :-( > I'll describe what I'm doing and maybe someone can point out where I'm going wrong or > else just tell me a completely different way of doing it. > > I first create a framework project that contains the code I want to wrap. When I > compile this it produces a framework bundle. > Following the documentation, I then create a wrapper project and add the header from my > framework (there's only one class in my framework), and the framework bundle itself > (oddly enough, it seems to compile fine without it). I'm reasonably sure my .jobs file > is okay as well. When this is compiled it produces a .dylib and some stuff in the > .build directory. I'm assuming that the only thing relevant to me in this directory is > the derived java source file which contains the native stubs (i don't know if this is > the right terminology). > I add this file and the framework bundle to the java application project that i want to > call the native code. This compiles fine, but when it runs it complains about either > not being able to find the framework or the dylib. So where do i put all this stuff? I > assume that the framework bundle goes in ~/Library/Frameworks/ but where does the dylib > go that the wrapper project produced? It would seem like a good place for it would be > inside the framework bundle, but moving it inside there doesn't seem to help. > > Any ideas? > > Cheers, > > Justin Armstrong Ok, I ended up (through trial and error) just putting the dylib in the same directory as the .app and this seems to work. This is not exactly ideal, though. Is there not some way of packaging this library and the .framework together into the .app bundle? Justin From hh at mdlink.de Mon Feb 22 13:55:56 1999 From: hh at mdlink.de (Helge Hess) Date: Thu Nov 3 14:41:51 2005 Subject: Solaris/Intel and PDO. References: <36C62DAD.66106905@powertrader.com> Message-ID: <36D1D26C.9337A2E6@mdlink.de> yonasj wrote: > > Does anybody know if PDO will work for Solaris/Intel. I've only seen PDO for Solaris/SPARC until now. You might want to try the GNUstep or libFoundation libraries which both run with Solaris/Intel. Helge -- MDlink online service center http://www.mdlink.de From kelleysoft at pressenter.com Mon Feb 22 14:37:56 1999 From: kelleysoft at pressenter.com (Greg Betzel) Date: Thu Nov 3 14:41:51 2005 Subject: Sharing ObjC headers In-Reply-To: <9902221837.AA08360@easynet.fr> References: Message-ID: Thanks to everyone for the responses. The solution, clearly, is to use @class. (Man, how did I overlook THAT one?? RTFM better next time, I guess :) Regards, Greg Betzel From Joe.Loda at usmail.mpct.com Tue Feb 23 06:08:24 1999 From: Joe.Loda at usmail.mpct.com (Loda, Joe) Date: Thu Nov 3 14:41:51 2005 Subject: Strange Compiler Warning Message-ID: Hello all, I'm getting a strange warning from the compiler (both OSE 4.2/NT and WO4/NT). The warning is "no previous prototype for _OBJC_MODULE_CONSTRUCTOR". I think it may have something to do with usage of @class and then subsequently defining the class (I tried this and couldn't reproduce it, however). Anyone have any ideas of what we may have done? Thanks, Joe --- Joe Loda, mpct Solutions Corporation, Chicago joe.loda@mpct.com (ASCII, MIME) From steve at estel.uindy.edu Tue Feb 23 06:34:31 1999 From: steve at estel.uindy.edu (Steve Spicklemire) Date: Thu Nov 3 14:41:51 2005 Subject: Simple shared libs? Message-ID: <199902231434.JAA28736@estel.uindy.edu> Hi MacOSXers... I'm trying to port some software that makes extensive use of shared libraries. Is there any simple way to fake the creation and use of .so files on DR2? Nothing I've tried seems to work. I'm thinking that bundles are sorta like shared libs.. so perhaps they could be used somehow? thanks, -steve From gerard at ina.fr Tue Feb 23 07:31:19 1999 From: gerard at ina.fr (=?iso-8859-1?Q?G=E9rard?= Iglesias) Date: Thu Nov 3 14:41:51 2005 Subject: Simple shared libs? References: <199902231434.JAA28736@estel.uindy.edu> Message-ID: <36D2C9C7.F9CD6263@ina.fr> Steve Spicklemire wrote: > > Hi MacOSXers... > > I'm trying to port some software that makes extensive use of shared > libraries. Is there any simple way to fake the creation and use of .so > files on DR2? Nothing I've tried seems to work. I'm thinking that > bundles are sorta like shared libs.. so perhaps they could be used > somehow? The good way to build shared lib is to make framework, bundles are also dynamic but for an other use. Sincerely. Gerard From nberch at db.lv Wed Feb 24 04:44:15 1999 From: nberch at db.lv (Nils Berzins) Date: Thu Nov 3 14:41:51 2005 Subject: Mach-O plugins... Message-ID: <199902241246.OAA15904@saulite> >> info from Metrowerks site... > > This update replaces all Mach-O plugins that were shipped in the Pre-release > folder on the CW Pro4 Tools CD. CodeWarrior Pro4 includes MacOS-Hosted plugins > for generating applications and libraries for MacOS X Server (Rhapsody) using > the CodeWarrior IDE. This is the easiest way to manage projects compared to > using traditional command-line compilers, like mwcc and gcc with makefiles. > You must be using CodeWarrior Pro 4 with the pre-release ("thrill seeker") > Mach-O plugins to use this update. These components are still alpha quality > and it is not recommended you attempt to use these for final, shipping code. > Hi, Has anyone here tried to use prerelease Mach-O plugins with Metrowerks CW4 ? I downloaded and installed them, but unfortunately I could not find any documentation describing how to use them. Installer doesn't seem to install any headers, libraries or frameworks. Only C/C++ compiler and Mach-O linker... Nils Berzins From jman at clubi.ie Wed Feb 24 13:08:30 1999 From: jman at clubi.ie (justin armstrong) Date: Thu Nov 3 14:41:51 2005 Subject: ObjC/Java bridge problems References: <36D1A85C.ADE43D8D@clubi.ie> Message-ID: <36D46A4E.A3EC02D7@clubi.ie> I've reduced my question down to one sentence: In MacOS X Server, what paths are searched at run time for dynamic libraries? From bbum at codefab.com Wed Feb 24 13:48:42 1999 From: bbum at codefab.com (Bill Bumgarner) Date: Thu Nov 3 14:41:51 2005 Subject: Linking Whackiness Re: ObjC/Java bridge problems References: <36D46A4E.A3EC02D7@clubi.ie> Message-ID: <36D473BA.85FC4773@mail.codefab.com> [I haven't confirmed this for dylibs-- but i'm pretty sure this remains true as a dylib is nothing more than a framework executable with a funky extension and no wrapper directory] None. The searching for dylibs only happens at link time... Wherever the dylib/framework is found, that is where it must be every time the executable is run. Actually, I lied. When building a framework, you can actually specify an "install name" (see -install_name on libtool). Basically, the install name is the full path to wherever the framework should be installed. By using the install name creatively, one can build entire trees of projects-- lots of frameworks and apps/tools-- someplace other than the final installation spot (see below). When a framework is built, the install name is actually encoded in the framework. As well, whenever a framework or other executable is linked, the full paths to all other frameworks/dylibs are included in the executable. You can see which frameworks are used and where they should be located by: Mac OS X: % otool -L executable Solaris: % ldd executable NT: right click teh executable, select Quick View and look at the import table. --- Why are the full paths encoded everywhere on Solaris and NT? To avoid serious risk of a security violation. If it were possible for the user to move and/or modify the framework or cause the app to link against a "fake" framework that dyna-loads the real thing, then it would be a relatively easy to task to violate the app in just about anyway one wanted. --- What is different about NT? NT does use a set of paths to search for DLLs on. It'll use the first one of any given name that it happens to find. This has advantages and disadvantages. Primary advantage is that you can shove DLLs anywhere and move 'em about freely-- as long as they are in the search path, they will be find. However, the disadvantages are numerous and have caused no end of hair pulling. I can't tell you how much time I or eithers I'm working with have lost because there was an old version of a DLL in the search path before the version under active development. Also, because the DLL must be in the search path, it means that development quality DLLs often end up in the search path and, if that DLL is used across multiple apps, may screw up other applications. Because the search path must have a set of directories, the DLL for a framework resides outside the framework-- typically in an Executables directory. This is a drag-- it means that it is impossible to consider a framework as a self contained resource. --- (see below) I have written something like OmniMake-- only a hell of a lot simpler-- that allows one to build standalone sets of frameworks and apps for testing/debugging purposes. It can also control information about install versions without necessarily having to adjust every single PB.project. It is a complete hacque, totally undocumented, and still requires a bit of tweakage to make work.... but has saved us a tremendous amount of time and is being used in production environments. If anyone wants to play, send me email. It requires WO 4.0 and the multiscript shell (see ftp://ftp.codefab.com/pub/unsupported/). b.bum justin armstrong wrote: > I've reduced my question down to one sentence: > In MacOS X Server, what paths are searched at run time for dynamic libraries? From jman at clubi.ie Wed Feb 24 17:45:59 1999 From: jman at clubi.ie (justin armstrong) Date: Thu Nov 3 14:41:51 2005 Subject: Linking Whackiness Re: ObjC/Java bridge problems References: <36D46A4E.A3EC02D7@clubi.ie> <36D473BA.85FC4773@mail.codefab.com> Message-ID: <36D4AB57.6DEEE0C@clubi.ie> Thanks to anyone thats helped me so far... Here's my situation now: I have created a (ObjC) framework that is living in my ~/Library/Frameworks directory I have a java wrapper project that takes this framework and produces a dylib. The default settings for a wrapper project install this in the /lib directory (At least, thats what it says in the PB inspector, but it doesn't actually put it there - so i just put a link to its real location in the /lib directory). Now i have a dylib and a framework. Do I need both of these or is the dylib supposed to contain all the stuff from the framework as well? I'm writing a Java application that I want to be able to use the code from the framework. I can add the framework to the app's project but is there some way i can let it know about the dylib as well ? Adding it to the libraries suitcase of PB appears to make no difference. Looking at the produced executable with otool shows that it knows the path of the framework but not the dylib. If i understand Bill right, the app won't find the dylib by searching certain paths, but instead relies on paths hardcoded into the executable. So how can i compile my application to know where to look for the dylib? And why is this not documented anywhere???? :-{ (my apologies if it is, please tell me where!) Cheers, Justin From jman at clubi.ie Wed Feb 24 17:53:45 1999 From: jman at clubi.ie (justin armstrong) Date: Thu Nov 3 14:41:51 2005 Subject: Linking Whackiness Re: ObjC/Java bridge problems References: <36D4AB57.6DEEE0C@clubi.ie> Message-ID: <36D4AD29.C84D5403@clubi.ie> > And why is this not documented anywhere???? :-{ > (my apologies if it is, please tell me where!) > other than the man pages ! From ghopson at novazen.com Wed Feb 24 18:05:16 1999 From: ghopson at novazen.com (Geoff Hopson) Date: Thu Nov 3 14:41:51 2005 Subject: Linking Whackiness Re: ObjC/Java bridge problems References: <36D4AB57.6DEEE0C@clubi.ie> Message-ID: <36D4AFDC.C3992F68@novazen.com> Have you tried explicitly including (with the -L directive) the libraries in the LIBS variable in the Makefile? LIBS = -L$(NEXT_ROOT)/LocalDeveloper/Libraries/libNZWrappedBase.so \ -L$(NEXT_ROOT)/LocalDeveloper/Libraries/libNZWrappedInputManager.so \ -L$(NEXT_ROOT)/LocalDeveloper/Libraries/libNZWrappedNotification.so \ -L$(NEXT_ROOT)/LocalDeveloper/Libraries/libNZWrappedEBPModel.so \ -L$(NEXT_ROOT)/LocalDeveloper/Libraries/libNZWrappedWebAppSupport.so ...or whatever your java-wrapped libraries are called... Geoff justin armstrong wrote: > > Thanks to anyone thats helped me so far... > Here's my situation now: > I have created a (ObjC) framework that is living in my ~/Library/Frameworks directory > > I have a java wrapper project that takes this framework and produces a dylib. The > default settings for a wrapper project install this in the /lib directory (At least, > thats what it says in the PB inspector, but it doesn't actually put it there - so i > just put a link to its real location in the /lib directory). > Now i have a dylib and a framework. Do I need both of these or is the dylib supposed > to contain all the stuff from the framework as well? > I'm writing a Java application that I want to be able to use the code from the > framework. I can add the framework to the app's project but is there some way i can > let it know about the dylib as well ? Adding it to the libraries suitcase of PB > appears to make no difference. Looking at the produced executable with otool shows > that it knows the path of the framework but not the dylib. > If i understand Bill right, the app won't find the dylib by searching certain paths, > but instead relies on paths hardcoded into the executable. So how can i compile my > application to know where to look for the dylib? > > And why is this not documented anywhere???? :-{ > (my apologies if it is, please tell me where!) > > Cheers, > > Justin From sanguish at digifix.com Wed Feb 24 20:55:21 1999 From: sanguish at digifix.com (Scott Anguish) Date: Thu Nov 3 14:41:51 2005 Subject: Stopping the default "new" document in Document-Based apps Message-ID: <199902250453.XAA24627@digifix.com> Has anyone else done this yet? Ideally we need a dwrite for each application... I worked around this by putting this in my Document class - init { static BOOL firstDocument = NO; if (!firstDocument) { firstDocument=YES; return nil; } return [super init]; } I'd much rather subclass NSDocumentController.. but for some reason I don't see how I'm supposed to do that (i.e. how do you set the controller to use???) Anyone know? (I'm almost positive I've done this before too... I just can't find this bit of Info) Thanks Scott From hunter at lastonepicked.com Wed Feb 24 21:16:48 1999 From: hunter at lastonepicked.com (Hunter Hillegas) Date: Thu Nov 3 14:41:51 2005 Subject: Where is MacOS X Server right now? Message-ID: <19990225051946928.AAA699@nts1.webhostingprovider.com@[207.71.222.134]> Hmmm... Does anyone else feel like Apple is about to break another promise? MOSXS still isn't shipping. I realize they technically have until Sunday to keep the promise, but please! Also, Select developers still don't have pricing information for the product... The last email on MOSXS from dev support was about 4 weeks ago and said they should have the information to use the following week. LAME! Hunter From cwolf at wolfware.com Wed Feb 24 21:37:29 1999 From: cwolf at wolfware.com (Christopher Wolf) Date: Thu Nov 3 14:41:51 2005 Subject: Stopping the default "new" document in Document-Based apps Message-ID: <199902250534.AAA04681@mail.rdu.bellsouth.net> In your Application delegate implement this method: - (BOOL) applicationShouldOpenUntitledFile: (NSApplication *)sender so that it returns no. This doesn't seem to be documented except in the header files. Chris > Has anyone else done this yet? Ideally we need a dwrite for each > application... > > I worked around this by putting this in my Document class > > - init > { > static BOOL firstDocument = NO; > > if (!firstDocument) > { > firstDocument=YES; > return nil; > } > return [super init]; > } > > > I'd much rather subclass NSDocumentController.. but for some reason > I don't see how I'm supposed to do that (i.e. how do you set the > controller to use???) Anyone know? (I'm almost positive I've done > this before too... I just can't find this bit of Info) > > Thanks > > Scott From sanguish at digifix.com Wed Feb 24 21:35:58 1999 From: sanguish at digifix.com (Scott Anguish) Date: Thu Nov 3 14:41:51 2005 Subject: Stopping the default "new" document in Document-Based apps In-Reply-To: <199902250534.AAA04681@mail.rdu.bellsouth.net> References: <199902250534.AAA04681@mail.rdu.bellsouth.net> Message-ID: <199902250536.AAA24820@digifix.com> Christopher Wolf wrote: > In your Application delegate implement this method: > > - (BOOL) applicationShouldOpenUntitledFile: (NSApplication *)sender > > so that it returns no. This doesn't seem to be documented except > in the header files. > Well THATS much easier.. :-) Any ideas on the NSDocumentController? > Chris > > > Has anyone else done this yet? Ideally we need a dwrite for each > > application... > > > > I worked around this by putting this in my Document class > > > > - init > > { > > static BOOL firstDocument = NO; > > > > if (!firstDocument) > > { > > firstDocument=YES; > > return nil; > > } > > return [super init]; > > } > > > > > > I'd much rather subclass NSDocumentController.. but for some reason > > I don't see how I'm supposed to do that (i.e. how do you set the > > controller to use???) Anyone know? (I'm almost positive I've done > > this before too... I just can't find this bit of Info) > > > > Thanks > > > > Scott From mpalluat at chez.com Wed Feb 24 23:53:21 1999 From: mpalluat at chez.com (Marc Palluat de Besset) Date: Thu Nov 3 14:41:51 2005 Subject: Stopping the default "new" document in Document-Based apps Message-ID: Subclassing NSDocumentController isn't recommended by Apple, but you can do it by defining a category on NSDocumentController and redefine the +sharedDocumentController class method as follow : +(id)sharedDocumentController { return [mySubclass sharedDocumentController]; } I hope this helps. Marc. From sanguish at digifix.com Wed Feb 24 23:58:05 1999 From: sanguish at digifix.com (Scott Anguish) Date: Thu Nov 3 14:41:51 2005 Subject: Stopping the default "new" document in Document-Based apps In-Reply-To: <199902250536.AAA24820@digifix.com> References: <199902250536.AAA24820@digifix.com> Message-ID: <199902250758.CAA25635@digifix.com> Scott Anguish wrote: > > Any ideas on the NSDocumentController? > OK.. so I've still not found this out.. but I did find that Yap seems to use an NSDocumentController subclass YapController. It seems to be instaniated in IB, but no connections are made to it... and no references are made to it either that I've been able to track down. I suppose that just instantiating an NSDocumentController could cause it to be the used one... But does it? Apple? Anyone? From gerard at ina.fr Thu Feb 25 00:07:44 1999 From: gerard at ina.fr (=?iso-8859-1?Q?G=E9rard?= Iglesias) Date: Thu Nov 3 14:41:51 2005 Subject: Stopping the default "new" document in Document-Based apps References: <199902250453.XAA24627@digifix.com> Message-ID: <36D504D0.B00E92D7@ina.fr> Scott Anguish wrote: > > Has anyone else done this yet? Ideally we need a dwrite for each > application... > > I worked around this by putting this in my Document class > > - init > { > static BOOL firstDocument = NO; > > if (!firstDocument) > { > firstDocument=YES; > return nil; > } > return [super init]; > } > > I'd much rather subclass NSDocumentController.. but for some reason > I don't see how I'm supposed to do that (i.e. how do you set the > controller to use???) Anyone know? (I'm almost positive I've done > this before too... I just can't find this bit of Info) > Two things: 1/ the delegate of the NSApp have to respond to - (BOOL)applicationShouldOpenUntitledFile:(NSApplication *)sender { return NO; } 2/ If you have to subclass NSDocumentController(I needed it a lot), you have to instantiate your subclass in the main nib, the system will find it: very strange isn't it? I found this by exploring the Draw.app sources and resources. I found nothing about this in the documentation. In fact, I don't like the design of the YB multi doc too much. It is based on the principe that each document mnage all its necessary windows, but it is not always true. For example we have a multi doc app which contains a lot of inspectors (View/Controller) that need to be synchronized with the current main doc and I had a lot of difficulty to have the thing together (it was the principal reason for what I subclassed NSDocumentController). In fact the currentDocument method can return nil, because it looks at the main window which can be nil in certain circumstances. Sincerely. Gerard. From saut at dptmaths.ens-cachan.fr Thu Feb 25 02:00:45 1999 From: saut at dptmaths.ens-cachan.fr (Olivier Saut) Date: Thu Nov 3 14:41:51 2005 Subject: Getting NSApplication arguments from the delegate Message-ID: <9902251100.AA452134@marid> Hello, I'd like to get the arguments my app receive, the default main from PB is : int main(int argc, const char *argv[]) { return NSApplicationMain(argc, argv); } How can I get these arguments from my app delegate? I can't find anything in the NSApplication documentation. Is it safe to call a YB app from the Terminal? (In fact what I am trying to do is to write a app to get the samba messages like WinPopUp does.) Thanks for your help, -- Olivier From Philippe.Robert at uptime.ch Thu Feb 25 02:02:39 1999 From: Philippe.Robert at uptime.ch (Philippe Robert) Date: Thu Nov 3 14:41:51 2005 Subject: Getting NSApplication arguments from the delegate In-Reply-To: <9902251100.AA452134@marid> References: <9902251100.AA452134@marid> Message-ID: <9902251003.AA27613@california.uptime.ch> You wrote: > Hello, > > I'd like to get the arguments my app receive, the default main from PB is : > > int main(int argc, const char *argv[]) { > return NSApplicationMain(argc, argv); > } > > > How can I get these arguments from my app delegate? I can't find anything > in the NSApplication documentation. > NSProcessInfo implements the methods arguments and environment. You can use these. 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 jjfeiler at diga.serv.net Thu Feb 25 06:07:30 1999 From: jjfeiler at diga.serv.net (John Jay Feiler) Date: Thu Nov 3 14:41:51 2005 Subject: Stopping the default "new" document in Document-Based apps In-Reply-To: <199902250758.CAA25635@digifix.com> References: <199902250758.CAA25635@digifix.com> Message-ID: <9902251407.AA12152@relief.com> A non-text attachment was scrubbed... Name: not available Type: text/enriched Size: 1513 bytes Desc: not available Url : /mailman/archive/macosx-dev/attachments/19990225/4a1e5ba4/attachment.bin From bbum at codefab.com Thu Feb 25 06:17:35 1999 From: bbum at codefab.com (Bill Bumgarner) Date: Thu Nov 3 14:41:51 2005 Subject: Getting NSApplication arguments from the delegate In-Reply-To: <9902251100.AA452134@marid> Message-ID: NSArray *appArgs = [[NSProcessInfo processInfo] arguments]; b.bum From wsanchez at apple.com Thu Feb 25 15:32:32 1999 From: wsanchez at apple.com (Wilfredo Sanchez) Date: Thu Nov 3 14:41:51 2005 Subject: Where is MacOS X Server right now? Message-ID: <199902252332.PAA09836@scv1.apple.com> | MOSXS still isn't shipping. We're working on it. Real hard. I promise. This isn't the right list for such discussions, though, so please call us lame elsewhere. -Fred From pierce at alumni.caltech.edu Thu Feb 25 15:54:20 1999 From: pierce at alumni.caltech.edu (Pierce T. Wetter III) Date: Thu Nov 3 14:41:51 2005 Subject: Where is MacOS X Server right now? In-Reply-To: <199902252332.PAA09836@scv1.apple.com> Message-ID: At 3:45 PM -0800 2/25/99, Wilfredo Sanchez wrote: > | MOSXS still isn't shipping. > > We're working on it. Real hard. I promise. > > This isn't the right list for such discussions, though, so please > call us lame elsewhere. > > -Fred Fred, you're not lame. You and all the Apple people are working really hard on Mac OS X. You've done quite a lot of work in a short amount of time. Three cheers! Its just that Yellow Box/WebObjects/EOF is SO COOL that all of us developer nerds are salivating at the mouth, and want it yesterday. For most, the last thing they saw was DR2. And DR2 was cool, and seemed pretty close. So no one can figure out why it isn't shipping yet. You know, and I know, and some others know its because you've added some other technologies to make OSX server more compelling then just being a development station. So in a way, you're in trouble because of your own coolness. 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 sanguish at digifix.com Thu Feb 25 20:52:14 1999 From: sanguish at digifix.com (Scott Anguish) Date: Thu Nov 3 14:41:51 2005 Subject: [resolved] Re: Stopping the default "new" document in Document-Based apps In-Reply-To: <9902251407.AA12152@relief.com> References: <9902251407.AA12152@relief.com> Message-ID: <199902260452.XAA00256@digifix.com> Thanks to all... and Apple, I've got my answer. The short answer is that Gerard Iglesias provided one correct answer, you can instantiate it in the nib. (this was my last guess on the list...) Another option is to do it in your application delegate object using applicationWillFinishLaunching:. This is probably what I'd opt for. However... The suggestion was made (by Apple DTS no less) that I might want to consider just making the New menu item point to my application delegate, and then have it call the appropriate NSDocumentController behaviour as required. I hadn't thought of that.. I was focused on the NSDocumentController aspect... Since what I need to do is force the user to specify a target directory before a new document is created (so that the document always has a disk location associated with it) this should work just fine. Thanks to all here, and to D Grits at Apple. From paulrs at lgs-systems.com Fri Feb 26 09:10:56 1999 From: paulrs at lgs-systems.com (Paul R. Summermatter) Date: Thu Nov 3 14:41:51 2005 Subject: Right mouse bug on NT Message-ID: <9902261710.AA00895@slab> RE: WOF 3.51 NT 4.0 Hey guys, I just found a nasty little bug which I hope is fixed in the latest release. On NT, right clicking on a text field nicely brings up a set of options to cut, copy and paste. Unfortunately, these options seem to ignore whether the field is disabled or not. You can actually cut information out of a disabled text field. This is UGLY! Has anyone tried this with the YB release included with WOF 4.0? I should be able to answer my own question by next week, but I would love to know now. 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 mike at lorax.com Sat Feb 27 09:41:08 1999 From: mike at lorax.com (mike) Date: Thu Nov 3 14:41:51 2005 Subject: Stopping the default "new" document in Document-Based apps Message-ID: <199902271837.KAA30373@boom.lorax.com> > It is based on the principe that each document mnage all its necessary > windows, but it is not always true. > I don't think this is quite right. It is based on the principle that a document might own one or more windows that it must manage. There may be other windows that are not managed by a specific document. These windows can still be managed by NSWindowControllers which are not owned by a document, if you wish. > For example we have a multi doc app which contains a lot of inspectors > (View/Controller) that need to be synchronized with the current main doc > and I had a lot of difficulty to have the thing together (it was the > principal reason for what I subclassed NSDocumentController). In fact > the currentDocument method can return nil, because it looks at the main > window which can be nil in certain circumstances. > I recommend looking at the Draw2/Sketch example (Draw2 in DR2, Sketch in OSXS). In general an NSDocument should own the NSWindowControllers for all the windows that belong exclusively to the document. NSWindowControllers for ancillary panels like inspectors should not be owned by a document. Draw2/Sketch shows how to use NSDocuments and NSWindowControllers in this way. DrawDocument creates and owns DrawWindowControllers which manage the main document window for each document. The InspectorController class is not created/owned by any document. It implements a sharedInstance API that allows the one InspectorController to be found and it observes what is going on in the app to keep its state up to date and in sync with whatever document is currently active. There are many ways that an inspector can figure out what to inspect. I don't believe it should be necessary to subclass NSDocumentController in order to implement inspectors reasonably. There is nothing wrong with subclassing NSDocumentController (although it is perhaps harder than it ought to be currently to get your subclass to be the one that creates the shared instance.) Mike Ferris From edream at mailbag.com Sun Feb 28 10:19:32 1999 From: edream at mailbag.com (Edward K. Ream) Date: Thu Nov 3 14:41:51 2005 Subject: MacOS X Server v MacOS X Message-ID: Hi all, This is intended to be a purely technical question, and I am not interested in any speculation, please. I don't understand the difference between MacOS X Server and MacOS X. Can someone explain the difference between them and tell why they should be different products? Why have a separate X Server at all? Edward ---------------------------------------------------- Edward K. Ream (608)-231-0766 edream@mailbag.com LEO: http://www.mailbag.com/users/edream/front.html Apple Computer, right down there with Mercedes Benz. ---------------------------------------------------- From dtm at vramp.net Sun Feb 28 10:38:20 1999 From: dtm at vramp.net (David T. McWherter) Date: Thu Nov 3 14:41:51 2005 Subject: MacOS X Server v MacOS X In-Reply-To: Message-ID: > I don't understand the difference between MacOS X Server and MacOS X. > Can someone explain the difference between them and tell why they > should be different products? Why have a separate X Server at all? Well, if they put Server in the name, then they can sell it for far more money (Behold what M$ does with NT and NT Server)... In addition, I think that MacOS X Server is supposed only a limited product...where you can play around with OpenSTEP and such while Apple finishes up on MacOS X, which will be more based on the MacOS - possibly removing the BSD server (although they promise POSIX compatability, although they say they won't ever bother to get certified), adding Carbon, etc. IIRC, MacOS X Server is supposed to only have a year's lifespan or so...(unless Apple changed their mind again). -David -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- David T. McWherter dtm@waterw.com From Malcolm_Crawford at plsys.co.uk Sun Feb 28 10:49:09 1999 From: Malcolm_Crawford at plsys.co.uk (mmalcolm crawford) Date: Thu Nov 3 14:41:51 2005 Subject: MacOS X Server v MacOS X In-Reply-To: References: Message-ID: <199902281849.SAA02027@jasmine.plsys.co.uk> Edward wrote: > I don't understand the difference between MacOS X Server > and MacOS X. > Primary differences: X will include the Carbon libraries so that MacOS apps can be readily recompiled to run on the platform. X will not include Display PostScript, it will be replaced by "Extended QuickDraw" with a light-weight window server (which will have particular benefits for YellowBox/NT, and no licensing issues). For X-Server applications this will mean that any using "PostScript wraps" (hand-crafted PostScript code sent directly to the window server) will not work. Most apps do not use pswraps, however, so in principle should work... but whilst most PostScript function calls will be mapped to the new drawing architecture, it is possible that there will be some problems so not all apps will work. For this reason developers are best advised to stick to simple pscommands and object-abstractions, such as NSBezier. It has also been suggested that there will be a "transparent" BlueBox; there is likely to be less Unix in evidence; it's possible there will be further UI refinements etc etc., but these all come under the heading "speculation", which you want us to avoid, and are not really relevant to this list (possibly with the exception of details of Unix-compatibility since this does affect some utilites)... ... all this said, Apple has indicated that X-Server is not a "V1.0-only" product, and that there will be further releases. It certainly makes sense to have workstation and server versions of a product, or at least distinct workstation and server products, since the optimal system configurations of the two differ (e.g. the former is likely to be optimised for responsiveness to user input, the latter to system throughput -- and may have additional applications and utilities). Best wishes, mmalc. From markm at mail.tyrell.com Sun Feb 28 10:51:07 1999 From: markm at mail.tyrell.com (Mark F. Murphy) Date: Thu Nov 3 14:41:51 2005 Subject: MacOS X Server v MacOS X In-Reply-To: Message-ID: At 10:27 AM -0800 2/28/99, Edward K. Ream wrote: >I don't understand the difference between MacOS X Server and MacOS X. >Can someone explain the difference between them and tell why they >should be different products? Why have a separate X Server at all? It's a little difficult to answer your question now because Apple doesn't really have any commitment for Mac OS X right now... there's just a few hints of their "plans". We know the big announced difference is what's called "Carbon". Carbon is a cleaned up sub-set of existing MacOS APIs which are designed to work in a pre-emptive environment as well as on the existing MacOS line. In fact according to rumor sites, Apple just released the CarbonLib/SDK to seeded developers under the existing MacOS. Currently, Mac OS X Server does not have the CarbonLib. Apple hasn't made any commitment to include the CarbonLib for Mac OS X Server now or in the future. Apple also revealed that their plan is to have a different graphics engine in Mac OS X than is in Mac OS X Server (which is currently display postscript). There are other differences as well which I won't spend the time to list here. The bottom line, in answer to your question, is that Apple has made forwarding looking statements which are not a commitment on their part. So to nail down an exact difference of what the two products will end up looking like once (and if) shipped is difficult to next to impossible. Though there will be some here who will try to convince you the future is secure... Apple has made promises... and will try to tell you the way things are "going" to be. I'd urge you to take such commentary from those people with a grain of salt and *not* make any plans based on it. It's fun to talk about what *might* happen... it becomes not fun when people have taken such fantasy chat as serious and spend their money and time on potential plans from Apple. Remember, we all know just about shit until the product actually ships. 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 anders at milkweed.com Sun Feb 28 11:36:22 1999 From: anders at milkweed.com (Anders Pytte) Date: Thu Nov 3 14:41:51 2005 Subject: MacOS X Server v MacOS X In-Reply-To: Message-ID: Ed, Perhaps others have missed the gist of your question, which I take to be why call it a Server rather than just a preliminary release of OSX. I would say the main reason is because the unix layer is exposed, making it inappropriate for consumers, but useful for server applications, and because it includes server software like apache, and WebObjects development environment, neither of which I expect to be wrapped with the consumer release (OSX). Anders. >Hi all, > >This is intended to be a purely technical question, >and I am not interested in any speculation, please. > >I don't understand the difference between MacOS X Server and MacOS X. >Can someone explain the difference between them and tell why they >should be different products? Why have a separate X Server at all? > >Edward > > >---------------------------------------------------- >Edward K. Ream (608)-231-0766 edream@mailbag.com >LEO: http://www.mailbag.com/users/edream/front.html >Apple Computer, right down there with Mercedes Benz. >---------------------------------------------------- ---------------------------------------------------------------- Anders Pytte Milkweed Software Ferguson Hill Voice: (802) 472-5142 Cabot VT 05647 Internet: anders@milkweed.com ---------------------------------------------------------------- From z.leo at iss.it Sun Feb 28 23:50:55 1999 From: z.leo at iss.it (Zanitti Leo) Date: Thu Nov 3 14:41:51 2005 Subject: Modal Panel? Message-ID: <3.0.5.32.19990301085055.007b6310@iss.it> My panel must run in modal mode for data entry: no other program activity mus be permitted. (it would be like NSRunAlertPanel) The panel have two fields for data entry with two formatters (Date and Number) and two buttons (Cancel and OK) How can I tell to the my panel to run modal mode Thanks :-) ;>) )^> %^) =^) :) >-) (:|) >:-( 8-) %^| Zanitti Leo :<( 8-> Via Capovilla n. 22 :-] :^< I-33010 Montenars (UD) :o) :-0 :-[ :=) E-Mail: z.leo@iss.it :<| ;-[ 8^| :^0 TEL +39 432 972966 :*) &^) ;=J 8-! :-/ :-I :-> :0) ;=| :>) :#( 3-) 8#( :-] 8*I