From bruce@trwlasd.com Wed Jan 29 17:01:01 1997 Return-Path: Received: from uucp11.netcom.com by omnigroup.com (NX5.67g/NX3.0M) id AA21267; Wed, 29 Jan 97 17:01:01 -0800 Received: from trwlasd.com by netcomsv.netcom.com with UUCP (8.6.12/SMI-4.1) id QAA25971; Wed, 29 Jan 1997 16:53:55 -0800 Received: from ewsdev13 by trwlasd.com with SMTP id AA07770 (5.65c/IDA-1.4.5 for ); Wed, 29 Jan 1997 16:21:09 -0800 Message-Id: <199701300021.AA07770@trwlasd.com> Received: by ewsdev13 (NX5.67d/NX3.0X) id AA02300; Wed, 29 Jan 97 16:21:08 -0800 Content-Type: text/plain Mime-Version: 1.0 (NeXT Mail 3.3 v118.2) Received: by NeXT.Mailer (1.118.2) From: Bruce McKenzie Date: Wed, 29 Jan 97 16:21:06 -0800 To: rhapsody-dev@omnigroup.com Subject: Traffic check... Reply-To: bruce@trwlasd.com This list seems awfully quiet. Have all possible questions been answered, and all topics discussed? :-) Bruce McKenzie DoublePup Enterprises, Inc. bruce@doublepup.com From jcr@idiom.com Sun Mar 16 18:55:37 1997 Return-Path: Received: from idiom.com by omnigroup.com (NX5.67f2/NX3.0M) id AA08138; Sun, 16 Mar 97 18:55:37 -0800 Received: (from jcr@localhost) by idiom.com (8.8.5/8.8.5) id SAA05809 for rhapsody-dev@omnigroup.com; Sun, 16 Mar 1997 18:55:21 -0800 (PST) From: "John C. Randolph" Message-Id: <199703170255.SAA05809@idiom.com> Subject: Introductions! To: rhapsody-dev@omnigroup.com Date: Sun, 16 Mar 1997 18:55:21 -0800 (PST) X-Mailer: ELM [version 2.4 PL24] Content-Type: text Hello all! This is my first message to this list. The NeXT crowd know me, for the most part, so for the Apple folks, here's my background in a nutshell: I started Macintosh hacking in 1984, and did some serious hardware bashing in 1984-1987, including developing the first real-time 30 frames/sec video digitizer for the Mac II. I was a Mac developer for five years alltogether, and I switched to NEXTSTEP in 1989. In the NeXT world, I've done software exclusively, and I have in-depth knowledge of how to hack the appkit, the runtime system, and the interface builder. I don't mind answering questions about nextstep, but when it comes to comparisons between the Mac and NeXT ways of doing things, I must warn you that I'm six years out of date on the Mac OS. On a nostalgic note, I was NeXT developer #343, and I was Macintosh developer #1187. And, ohe thing I'll ask you to just take my word on: once you start hacking with Objects; real, xerox-parc-style objects, with dynamic binding, and a rich set of prebuilt cmponents, you'll never miss the old way one bit! -jcr From pcoffman@pobox1.stanford.edu Sun Mar 16 21:37:32 1997 Return-Path: Received: from mailhub.Stanford.EDU by omnigroup.com (NX5.67f2/NX3.0M) id AA10401; Sun, 16 Mar 97 21:37:32 -0800 Received: from [171.64.185.42] (infosys.Stanford.EDU [171.64.185.42]) by mailhub.Stanford.EDU (8.8.5/8.8.5/L) with SMTP id VAA16447 for ; Sun, 16 Mar 1997 21:37:12 -0800 (PST) Message-Id: <199703170537.VAA16447@mailhub.Stanford.EDU> Subject: testing Date: Sun, 16 Mar 97 21:37:14 -0800 X-Sender: pcoffman@pobox1.stanford.edu X-Mailer: Claris Emailer 2.0 x52, February 26, 1997 From: "Patrick L. Coffman" To: "Rhapsody Developer List" Mime-Version: 1.0 Content-Type: text/plain; charset="US-ASCII" sorry, i needed to try this... please forgive me. From abridge@wheel.dcn.davis.ca.us Sun Mar 16 22:28:32 1997 Return-Path: Received: from wheel.dcn.davis.ca.us by omnigroup.com (NX5.67f2/NX3.0M) id AA11078; Sun, 16 Mar 97 22:28:32 -0800 Received: from [168.150.253.54] (dcn54.dcn.davis.ca.us [168.150.253.54]) by wheel.dcn.davis.ca.us (8.8.5/DCN8.8.0) with SMTP id WAA14037 for ; Sun, 16 Mar 1997 22:28:13 -0800 (PST) Message-Id: <199703170628.WAA14037@wheel.dcn.davis.ca.us> Subject: Moving to Rhapsody from Smalltalk Date: Sun, 16 Mar 97 22:28:13 -0800 X-Mailer: Claris Emailer 2.0 x49, February 10, 1997 From: Adam Bridge To: Mime-Version: 1.0 Content-Type: text/plain; charset="US-ASCII" I'm currently developing a piece of software entirely in QKS' SmalltalkAgents with the plan to move over to the Rhapsody environment as soon as it is released. Can someone familier with Smalltalk discuss the changes in orientation I'll need to make in order to deal with OpenStep and Objective-C. I've seen code fragments and read some of the basic tutorials but that's about it for Objective-C. I understand I'll have to learn new frameworks but they all seem well documented. I know I'll miss the intimate development environment that Smalltalk offers. But other insights, and perhaps comments by those who have come to the NeXT envrionment from a Smalltalk background would be greatly appreciated. Best regards, Adam Bridge From sanguish@digifix.com Sun Mar 16 22:54:03 1997 Return-Path: Received: from digifix.digifix.com by omnigroup.com (NX5.67f2/NX3.0M) id AA11596; Sun, 16 Mar 97 22:54:03 -0800 Received: (from sanguish@localhost) by digifix.com (8.8.5/8.8.5) id BAA05577 for rhapsody-dev@omnigroup.com; Mon, 17 Mar 1997 01:53:40 -0500 (EST) Message-Id: <199703170653.BAA05577@digifix.com> Content-Type: text/plain Mime-Version: 1.0 (NeXT Mail 4.1mach v148) X-Nextstep-Mailer: Mail 4.1mach (Enhance 2.0b5) Received: by NeXT.Mailer (1.148) From: Scott Anguish Date: Mon, 17 Mar 97 01:53:38 -0500 To: rhapsody-dev@omnigroup.com Subject: Request for Input - Evangelizing an OpenStep/Rhapsody solution for Scanner input Reply-To: sanguish@digifix.com PC and Macintosh have TWAIN, but having a brief lookat the 344 pages of documentation on Twain, I'm thinking THERE MUST BE AN EASIER WAY! Since TWAIN isn't on OpenStep, and since it looks like an object oriented solution would be much nicer anyways, I'd like to get some input on what would be required by developers before trying to find a place for this within NeXT or Apple. Right now, companies like interpersonal-computing and GSCorp have scanning software on NEXTSTEP... I'm looking forward, more to future situation when there might be a case for shipping a scanner 'driver' that can just be dropped into /LocalLibrary/InputManagers (probably not the right place for it) and made available to apps. There seems to be two different modes for twain. with UI, or without a UI.. There needs to be ways to query the scanner about what it is capable of and returning an NSDictionary seems natural here, returning pre-defined constants where appropriate and NSNumbers where real values are required. You'd need to be able to specify settings to get data in, specifying things like image size, resolution, some color correction information, some special features perhaps... definately needs to be workable across multiple types of devices (scanners, cameras, video-capture etc..) I'm also wondering about the needs in this for stuff like an NSImageMapEnumerator, an easy way to access individual pixels and rows of pixels... I've got some other ideas cooking in all this, but was wondering if there was any interest, or if someone else (inside Apple/NeXT is already working on this) I'd also REALLY like to see AppKit support for .gif and .png formats natively... From gad@eclipse.its.rpi.edu Mon Mar 17 04:37:32 1997 Return-Path: Received: from mail1.its.rpi.edu by omnigroup.com (NX5.67f2/NX3.0M) id AA15708; Mon, 17 Mar 97 04:37:32 -0800 Received: from eclipse.its.rpi.edu (gad@eclipse.its.rpi.edu [128.113.24.33]) by mail1.its.rpi.edu (8.8.5/8.8.5) with SMTP id HAA42822; Mon, 17 Mar 1997 07:37:28 -0500 Received: by eclipse.its.rpi.edu (NX5.67e/NX3.0M) id AA12519; Mon, 17 Mar 97 07:37:26 -0500 Message-Id: <9703171237.AA12519@eclipse.its.rpi.edu> Mime-Version: 1.0 (NeXT Mail 3.3 v118.2) Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Received: by NeXT.Mailer (1.118.2) From: Garance A Drosehn Date: Mon, 17 Mar 97 07:37:25 -0500 To: sanguish@digifix.com Subject: Re: Request for Input - Evangelizing an OpenStep/Rhapsody solution for Cc: Multiple recipients of list References: <199703170653.BAA05577@digifix.com> > I'd also REALLY like to see AppKit support for .gif and .png > formats natively... I don't care much about GIF support, but I'd certainly like to see .png support native in Rhapsody. One of the biggest attractions of PNG to me is that the format includes gamma-level information in it. Anyone who deals with images on both Macs (or PC's) and NeXTSTEP knows what an irritation it is that when trying to fudge the gamma levels. I'd think that a single system trying to be both OpenStep and MacOS would have to support some image type which explicitly understands gamma settings. --- Garance Alistair Drosehn = gad@eclipse.its.rpi.edu Senior Systems Programmer (MIME & NeXTmail capable) Rensselaer Polytechnic Institute; Troy NY USA From msb@plexare.com Mon Mar 17 05:49:15 1997 Return-Path: Received: from empire.is.com by omnigroup.com (NX5.67f2/NX3.0M) id AA16984; Mon, 17 Mar 97 05:49:15 -0800 Received: from mspboss.is.com by empire.is.com; Mon, 17 Mar 1997 07:49:09 -0600 Received: by mspboss.is.com with UUCP (8.6.12/IS-Mailhost-1.0.1/plex_net33) id HAA27754; Mon, 17 Mar 1997 07:48:37 -0600 Received: from imladris by plexare.com with SMTP (NX5.67f2/Plexare-1.0.2) id AA03129; Mon, 17 Mar 97 07:54:52 -0600 Message-Id: <9703171354.AA03129@plexare.com> Received: by imladris.plexare.com (NX5.67g/NX3.0X) id AA00459; Mon, 17 Mar 97 07:54:55 -0600 Mime-Version: 1.0 (NeXT Mail 4.1mach v148) Content-Type: text/enriched; charset=us-ascii Content-Transfer-Encoding: 7bit Content-Disposition: inline Received: by NeXT.Mailer (1.148) From: "Michael S. Barthelemy" Date: Mon, 17 Mar 97 07:54:54 -0600 To: gad@eclipse.its.rpi.edu Subject: Re: Request for Input - Evangelizing an OpenStep/Rhapsody solution Cc: Multiple recipients of list Reply-To: msb@is.com References: <9703171237.AA12519@eclipse.its.rpi.edu> TIFF supports having gamma curves within it's format - just most people don't use them. Mike Barthelemy msb@plexare.com Begin forwarded message: Date: Mon, 17 Mar 97 04:39:53 -0800 Reply-To: gad@eclipse.its.rpi.edu Sender: rhapsody-dev@omnigroup.com From: Garance A Drosehn < To: Multiple recipients of list < Subject: Re: Request for Input - Evangelizing an OpenStep/Rhapsody solution > I'd also REALLY like to see AppKit support for .gif and .png > formats natively... I don't care much about GIF support, but I'd certainly like to see .png support native in Rhapsody. One of the biggest attractions of PNG to me is that the format includes gamma-level information in it. Anyone who deals with images on both Macs (or PC's) and NeXTSTEP knows what an irritation it is that when trying to fudge the gamma levels. I'd think that a single system trying to be both OpenStep and MacOS would have to support some image type which explicitly understands gamma settings. --- Garance Alistair Drosehn = gad@eclipse.its.rpi.edu Senior Systems Programmer (MIME & NeXTmail capable) Rensselaer Polytechnic Institute; Troy NY USA From markm@mail.tyrell.com Mon Mar 17 14:52:14 1997 Return-Path: Received: from apex.tyrell.com by omnigroup.com (NX5.67f2/NX3.0M) id AA29396; Mon, 17 Mar 97 14:52:14 -0800 Received: from [206.170.253.84] (cdungeon.tyrell.com [206.170.253.84]) by mail.tyrell.com (8.6.12/8.6.12-MT2.2) with ESMTP id OAA23369 for ; Mon, 17 Mar 1997 14:51:01 PST (-0800) Message-Id: Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Date: Mon, 17 Mar 1997 14:52:29 -0700 To: rhapsody-dev@omnigroup.com From: "Mark F. Murphy" Subject: Hello Well... I've been subscribed for several hours and not one message! So I might as well say hello! I'd like to thank Omni Development and William Shipley for this list and their desire to help Mac developers with this transition we face together of both our worlds! I'm very excited about the opportunity to develop with an advanced environment like OpenStep.... something the Mac world has needed for quite some time. I think Rhapsody on Mac PPC hardware is an excellent opportunity to create software that should blow the socks off of the Windows environment. I also look forward to learning a lot from this list. Hopefully this list will steer clear of some of the clutter that is on semper Fi and concentrate on pure Rhapsody development. Sincerely, Mark F. Murphy --------------------------------------------------------------------------- Mark F. Murphy, Director Software Development, Tyrell (markm@tyrell.com) Sysop, The Desktop BBS (Macintosh) 714-491-1003 --------------------------------------------------------------------------- Member of Families Against Internet Censorship: rainbow.rmi.net/~fagin/faic From mpinkert@lennon.cc.gatech.edu Mon Mar 17 17:00:33 1997 Return-Path: Received: from burdell.cc.gatech.edu by omnigroup.com (NX5.67f2/NX3.0M) id AA01774; Mon, 17 Mar 97 17:00:33 -0800 Received: from lennon.cc.gatech.edu (root@lennon.cc.gatech.edu [130.207.9.20]) by burdell.cc.gatech.edu (8.8.4/8.6.9) with ESMTP id TAA06865 for ; Mon, 17 Mar 1997 19:52:58 -0500 (EST) Received: from [130.207.118.117] (fire.cc.gatech.edu [130.207.118.117]) by lennon.cc.gatech.edu (8.8.4/8.6.9) with SMTP id TAA02972 for ; Mon, 17 Mar 1997 19:52:57 -0500 (EST) Message-Id: <199703180052.TAA02972@lennon.cc.gatech.edu> Subject: OpenStep documentation/Rhapsody Date: Mon, 17 Mar 97 19:51:13 -0500 X-Sender: mpinkert@lennon.cc.gatech.edu X-Mailer: Claris Emailer 2.0 x52, February 26, 1997 From: Mike Pinkerton To: "rhapsody-dev" Mime-Version: 1.0 Content-Type: text/plain; charset="US-ASCII" Hi folks If you're looking for documentation to get started with, check out: http://www.next.com/Pubs/Documents/Download/apple.html -- Mike Pinkerton mpinkert@cc.gatech.edu http://www.cc.gatech.edu/~mpinkert Cyberdog: On the Internet, no one knows you're an OpenDoc part From don@misckit.com Mon Mar 17 17:09:04 1997 Return-Path: Received: from misckit.com by omnigroup.com (NX5.67f2/NX3.0M) id AA01937; Mon, 17 Mar 97 17:09:04 -0800 Received: from brain by misckit.com (NX5.67e/NX3.0M) id AA24863; Mon, 17 Mar 97 18:05:34 -0700 Message-Id: <9703180105.AA24863@misckit.com> Received: by brain.misckit.com (NX5.67g/NX3.0X) id AA00685; Mon, 17 Mar 97 18:05:36 -0700 Content-Type: text/plain Mime-Version: 1.0 (NeXT Mail 4.1mach v148) In-Reply-To: X-Nextstep-Mailer: Mail 4.1mach (Enhance 2.0b5) Received: by NeXT.Mailer (1.148.RR) From: Donald A. Yacktman Date: Mon, 17 Mar 97 18:05:35 -0700 To: rhapsody-dev@omnigroup.com Subject: Drawing via bit twiddling and Re: Hello Reply-To: don@globalobjects.com References: Approved: don@misckit.com X-Warners: Yakko, Wakko, and Dot On Mon, 17 Mar 97, "Mark F. Murphy" wrote: > So I might as well say hello! Hi back! :-) > I think Rhapsody on Mac PPC hardware is an excellent opportunity to create > software that should blow the socks off of the Windows environment. Me too, and I certainly hope to be one of the ones doing it...or at the very least be someone creating enabling tools for others out there. To start of some real discussion, Mark mentioned to me in private email that he wants to discuss drawing to direct bitmaps under OPENSTEP. To stir things up, here's some food for thought. 1. Interceptor is not a panacea. In fact, if you aren't clever enough, DPS will outperform your direct-to-screen routines. And it isn't as idiot-proof as DPS, so it is easy to get bit--badly. 2. The same techniques required to program to the Interceptor interface will actually improve performance when you go through DPS! In fact, IMHO it is probably a good idea to program via DPS, debug everything, and then sneak the Interceptor support in after. So I think probably the best thing is to look at how you'd do this going through DPS. Two apps I've worked on are illustrative of the technique, and I strongly recommend that those who are interested in the discussion look at them--they come with source available. The first is the Apple //e emulator. We actually got this thing to emulate an Apple //e--including SOUND--on a 68040 NeXTStation Color and it worked faster than the original //e! That, to me at least, was a nifty accomplishment. It was neat to run frogger and have it work with sound and everything. Anyway, the point is that a LOT of fast performance was squeezed out of those color slabs and since source is available, it is worth figuring out what we did. It IS arcane... ftp://ftp.thoughtport.com/pub/next/misckit/working/apple2/emulator/ The second example would be the Lyapunov.app that I wrote--a fractal generator. It was not too concerned with screen performance since the time to calculate the fractal is at least an order of magnitude more than the drawing time. So it is a REALLY simple example of direct pixel twiddling. In that one, I laid out a bitmapped buffer in true color and let DPS convert it to a buffer that matches the screen. (ie, it isn't tuned to anything but a 24-bit display). Look for the sources at: ftp://next-ftp.peak.org/pub/next/sources/graphics/ Another resource would be, on 3.3 systems: /NextLibrary/Documentation/NextDev/Concepts/Performance/C_ImprovingDrawingPerformance.rtf /NextLibrary/Documentation/NextDev/ReleaseNotes/WindowServer.rtf For those without a NEXTSTEP 3.3 machine, access this doc via: http://www.misckit.com/Drawing/perf_note.html http://www.misckit.com/Drawing/ws_note.html (Converted to HTML from NeXT RTF by LatinByrd, one heck of a cool app...and then watch OmniWeb churn away at rendering these stinkers. :-) ) The first one is a MUST READ, and is still useful for OPENSTEP developers to know. The second discusses a performance enhancement in the PostScript "image" operator for 68040's which is important--that's a technique you need to use with Interceptor for the best performance. (Search for "68040".) Look for notes on NXDrawBitmap() as well. Also, if you're splatting bitmaps to the screen via DPS, look at using _retained_ windows instead of _buffered_ windows. Another place to look for code is in the old (and I do mean OLD) GameKit sources. I've done a lot of work since these were released, but the new stuff isn't useable and hence not releaseable yet. And it isn't OPENSTEP ready, so I really need to rework it before I re-release it. The old stuff DOES show how to make use of the NXImage object as an offscreen buffer to a retained window. Although I didn't do all the arcane optimizations, the most powerful optimization was done: selective buffer flushing. DPS coalesces all buffer changes into _one_ large rect when doing a buffer flush. I coalesce into multiple smaller rects. The tuning comes in deciding when to coalesce and when to create a new rect, and the optimal value varies per game. But even the defaults give a significant improvement in performance. The GameKit is at: ftp://ftp.thoughtport.com/pub/next/misckit/working/gamekit/ Most of what is there is now outdated because I've rewritten most of the kit to make it better--though most of the concepts you'll see will still be the same. This stuff gets terribly arcane really fast...but those willing to dig and experiment WILL be rewarded! That's a good start...and plenty to chew. I'll see if I can get a better explanation or a tutorial written to help beginners actually try some of this stuff out...right now I've got to get back to work. :-) --- Later, -Don Yacktman don@misckit.com My home page From markm@mail.tyrell.com Mon Mar 17 17:18:25 1997 Return-Path: Received: from apex.tyrell.com by omnigroup.com (NX5.67f2/NX3.0M) id AA02185; Mon, 17 Mar 97 17:18:25 -0800 Received: from [206.170.253.84] (cdungeon.tyrell.com [206.170.253.84]) by mail.tyrell.com (8.6.12/8.6.12-MT2.2) with ESMTP id RAA24124 for ; Mon, 17 Mar 1997 17:17:15 PST (-0800) Message-Id: In-Reply-To: <199703180052.TAA02972@lennon.cc.gatech.edu> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Date: Mon, 17 Mar 1997 17:18:42 -0700 To: rhapsody-dev@omnigroup.com From: "Mark F. Murphy" Subject: Drawing Directly To Screen I've been doing game ports for the last several years... so I'm very interested in knowing how to access the screen directly or via a good set of classes. On the Mac, I basically use a PixMap to get info about the port (color table, size, etc). But mostly... a routine called CopyBits is used to copy one PixMap to another. For those unfamilar, CopyBits not only copies from one port to the next... it can reshape... resize... use masks... use transfers modes... dither... transparency, etc. So as you can tell, it does quite a bit. For most things, CopyBits is fine (on a low level)... but at en even lower level, one might write directly into memory described by the PixMap for speed consideration. Also... there's some concepts in QuickDraw on the Mac that come in handy... like Regions (a non-rectangular description of an area). Regions can come in many different shapes and size (donuts, etc). I haven't looked at what QuickDraw GX offered (since it wasn't bundled with the MacOS)... but some say it was the next step up from what Color QuickDraw offered on the Mac. What way will we be able to access graphic devices... and still be able to address multiple monitors and keep cross platform in mind (i.e. the target platform might be Rhapsody or Win NT/95). I'm all ears! mark --------------------------------------------------------------------------- Mark F. Murphy, Director Software Development, Tyrell (markm@tyrell.com) Sysop, The Desktop BBS (Macintosh) 714-491-1003 --------------------------------------------------------------------------- Member of Families Against Internet Censorship: rainbow.rmi.net/~fagin/faic From Robert.Kieffer@Eng.Sun.COM Mon Mar 17 17:32:18 1997 Return-Path: Received: from venus.Sun.COM by omnigroup.com (NX5.67f2/NX3.0M) id AA02483; Mon, 17 Mar 97 17:32:18 -0800 Received: from Eng.Sun.COM ([129.146.1.25]) by venus.Sun.COM (SMI-8.6/mail.byaddr) with SMTP id RAA24361 for ; Mon, 17 Mar 1997 17:32:06 -0800 Received: from litehouse.eng.sun.com by Eng.Sun.COM (SMI-8.6/SMI-5.3) id RAA09759; Mon, 17 Mar 1997 17:32:04 -0800 Received: from lhbodie.Eng.Sun.Com by litehouse.eng.sun.com (SMI-8.6/SMI-SVR4) id RAA10645; Mon, 17 Mar 1997 17:32:04 -0800 Message-Id: <199703180132.RAA10645@litehouse.eng.sun.com> Received: by lhbodie.Eng.Sun.Com (NX5.67g/NX3.0X) id AA00616; Mon, 17 Mar 97 17:32:54 -0800 Mime-Version: 1.0 (NeXT Mail 4.1mach v148) Content-Type: text/enriched; charset=us-ascii Content-Transfer-Encoding: 7bit Content-Disposition: inline X-Nextstep-Mailer: Mail 4.0 (Enhance 2.0b4) Received: by NeXT.Mailer (1.148) From: Robert Kieffer Date: Mon, 17 Mar 97 17:32:52 -0800 To: rhapsody-dev@omnigroup.com Subject: Introduction for Robert Kieffer Reply-To: Robert.Kieffer@Eng.Sun.COM First of all, I'd like to say hi to all the old NeXTSteppers out there ("Hi") and introduce myself to the Apple half of this group: My name is Robert Kieffer. I'm a 7-year veteran NeXTStep developer. NeXTStep was my introduction to OO-programming so I have a fairly Objective-C/NeXTStep/OpenStep centric view of the world. I've been working at Lighthouse Design (which I'll assume most of you are familiar with) for the last 2-1/2 years and am one of the three principle developers on JavaPlan, our newly released OO-design tool. Also at Lighthouse, I worked on Quantrix 2.0 (the Lotus Improv clone we developed, an N-dimensional spreadsheet application). Prior to Lighthouse I cofounded a company called Alodar Systems with a group of coworkers from TRW Aerospace, where I had my first introduction to NeXT. While at Alodar I wrote an image archive management program called image Curator and at TRW I did a variety of R & D work involving image processing. 'nuff said on that. :-) My first couple of contributions to this group will be suggestions on ways I feel NeXTStep (soon to be Rhapsody) could better support 3rd party developers. I'll post these in follow-up emails so that the Subject line will be something reasonable. Cheers, and I look forward to talking with all of you. --- Robert Kieffer Phone: (415)524-3238 Lighthouse Design, Ltd. Fax: (415)570-7787 2929 Campus Drive, Suite 101 Email: kieffer@litehouse.eng.sun.com San Mateo, CA 94403 WWW: http://www.lighthouse.com/~kieffer From mpinkert@lennon.cc.gatech.edu Mon Mar 17 19:19:42 1997 Return-Path: Received: from burdell.cc.gatech.edu by omnigroup.com (NX5.67f2/NX3.0M) id AA04367; Mon, 17 Mar 97 19:19:42 -0800 Received: from lennon.cc.gatech.edu (root@lennon.cc.gatech.edu [130.207.9.20]) by burdell.cc.gatech.edu (8.8.4/8.6.9) with ESMTP id WAA18298 for ; Mon, 17 Mar 1997 22:19:34 -0500 (EST) Received: from [130.207.118.117] (fire.cc.gatech.edu [130.207.118.117]) by lennon.cc.gatech.edu (8.8.4/8.6.9) with SMTP id WAA09526 for ; Mon, 17 Mar 1997 22:19:32 -0500 (EST) Message-Id: <199703180319.WAA09526@lennon.cc.gatech.edu> Subject: NextStep vs OpenStep Date: Mon, 17 Mar 97 22:17:46 -0500 X-Sender: mpinkert@lennon.cc.gatech.edu X-Mailer: Claris Emailer 2.0 x52, February 26, 1997 From: Mike Pinkerton To: "rhapsody-dev" Mime-Version: 1.0 Content-Type: text/plain; charset="US-ASCII" Newbie question.... I see people throwing around references to NextStep and OpenStep as if they are worlds apart. What is the difference (if any) and which one is Apple using in Rhapsody? Thanks -- Mike Pinkerton mpinkert@cc.gatech.edu http://www.cc.gatech.edu/~mpinkert Cyberdog: On the Internet, no one knows you're an OpenDoc part From wjs Mon Mar 17 19:32:04 1997 Return-Path: Received: from foom by omnigroup.com (NX5.67f2/NX3.0M) id AA04726; Mon, 17 Mar 97 19:32:04 -0800 Message-Id: <9703180332.AA04726@omnigroup.com> Received: by foom.omnigroup.com (NX5.67g/NX3.0X) id AA06148; Mon, 17 Mar 97 19:32:03 -0800 Content-Type: text/plain Mime-Version: 1.0 (NeXT Mail 4.1mach v148) X-Nextstep-Mailer: Mail 3.3 (Enhance 1.3) Received: by NeXT.Mailer (1.148) From: William Shipley Date: Mon, 17 Mar 97 19:32:02 -0800 To: rhapsody-dev Subject: RE: Moving to Rhapsody from Smalltalk Reply-To: wjs@omnigroup.com X-Image-Url: ftp://ftp.omnigroup.com/pub/Images/People/wjs.tiff X-Organization: Omni Development, Inc. It's been a while since I've programmed SmallTalk, but I thought I'd give you some pointers anyways, and let someone more knowledgable fill in the gaps I leave. What you lose: - In Objective-C, you'll have to get used to other types besides objects. That is, we have all the primitive types that C does -- ints, floats, structs, etc. In SmallTalk, these types are represented as true objects, to which you can send messages. In Objective-C, you have to use C-syntax to deal with these types. (So, you can't add methods to an integer. Bummer.) Under OpenStep (and Java) they have classes which are just covers for the primitive types (NSValues in OpenStep, java.lang.Numbers in Java), so you can pass around, say, ints as first-class objects by wrapping them in an NSNumber object first. This also lets you add methods to ints. The disadvantage (compared to SmallTalk) is there is no compiler support for this, so it isn't as efficient (to my knowledge) to add two NSNumbers in ObjC as it is to add two ints in SmallTalk. - Similarly, in Objective-C we still have functions, as well as methods. Sometimes you'll have to call these -- hopefully you'll never write any yourself. If you stay within OpenStep, even calling functions is pretty rare, but it's still occasionally a yucky fact of life. - Objective-C doesn't have lambda functions, so you can't pass in code as a parameter to a method. It's been so darn long I can't even remember how you use lambda functions in SmallTalk, so I can't really say much about this one. - Objective-C doesn't have a garbage collector. Some of us think this is a good thing, some a bad. I don't wish to get into a religious war about it. In OpenStep, objects are retain-counted, and it's up to the programmer to make sure she retains objects that she wants to stick around. This is second nature for me now, and I personally much prefer it over Java's garbage collection, because it allows me to decide when and how I'm going to allocate and deallocate memory (and thus do optimizations), but some people will argue that this kind of low-level hacking has no place in today's high-level programs. (To those people, I'll point them at any Java program, which runs slower than a snail on downers.) What you gain: You'll pretty much have to read one of the tutorials on ObjC to learn what's cool about it. I posted some of the things I like to semper-fi a while ago, and there's also a tutorial on the Stepwise OpenStep site. Categories are darn nice. Being compatible with C is pretty nice in a lot of cases. It's very easy to tune ObjC in the places where speed is important, since you can always drop into straight-C or some hybrid. But, basically, ObjC and SmallTalk are an awfully lot alike. By the way, ProjectBuilder 4.0 reminds me of nothing so much as the SmallTalk development environment I used in college. It's a little different here and there, but in general everyone I know really loves it. (It even has integrated most of the common emacs keys, including incremental search and split window. It's just darn cool.) -Wil From mark@oaai.com Mon Mar 17 19:36:00 1997 Return-Path: Received: from www.oaai.com by omnigroup.com (NX5.67f2/NX3.0M) id AA04884; Mon, 17 Mar 97 19:36:00 -0800 Received: from sapphire ([199.166.28.100]) by sapphire.oaai.com (post.office MTA v2.0 0813 ID# 0-0U10) with ESMTP id AAA406; Mon, 17 Mar 1997 22:34:34 -0500 From: mark@oaai.com (Mark Onyschuk) To: Cc: Subject: Re: NextStep vs OpenStep Date: Mon, 17 Mar 1997 22:34:33 -0500 X-Msmail-Priority: Normal X-Priority: 3 X-Mailer: Microsoft Internet Mail 4.70.1155 Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Message-Id: <19970318033434363.AAA406@sapphire> NextStep is a precursor to OpenStep. The former is a NeXT-only technology while the latter has been implemented by both NeXT and Sun, and is currently being implemented by a collection of talented folks as part of the GNU initiative. OpenStep, being more modern than NextStep, improves on the original in several areas: * it provides a standard mechanism for reclaiming objects once they are no longer in use, a simple form of GC. * it provides a more comprehensive layer of objects which insulates the developer from platform-specific coding * it provides fundamental support for UNICODE strings. * it features an improved distributed object messaging framework To sum it up, it's NextStep with all of the ugly bits smoothed over. In my experience, the typical NextStep framework is reduced in size by a third when it's ported to OpenStep - the new framework does more and does it more cleanly. Regards, Mark --- M. Onyschuk and Associates Inc. Software Systems for the U.S. and Canadian Financial Industries. 15 La Rose Ave. Suite 702, Weston Ontario M9P 1A7 (416)241-3076 From dke@san.rr.com Mon Mar 17 19:37:51 1997 Return-Path: Received: from san.rr.com by omnigroup.com (NX5.67f2/NX3.0M) id AA05040; Mon, 17 Mar 97 19:37:51 -0800 Received: (from uucp@localhost) by mail.san.rr.com (8.7.6/8.7.3) id TAA06388; Mon, 17 Mar 1997 19:37:10 -0800 (PST) Message-Id: <199703180337.TAA06388@mail.san.rr.com> Received: from dt6h2n5a.san.rr.com(204.210.36.90) by mail via smap (V1.3) id tmp006331; Mon Mar 17 19:36:49 1997 Subject: Re: NextStep vs OpenStep Date: Mon, 17 Mar 97 19:47:28 -0800 X-Mailer: Claris Emailer 2.0 x49, February 10, 1997 From: "David K. Every" To: , "Multiple recipients of list" Mime-Version: 1.0 Content-Type: text/plain; charset="US-ASCII" >Newbie question.... > >I see people throwing around references to NextStep and OpenStep as if >they are worlds apart. What is the difference (if any) and which one is >Apple using in Rhapsody? My understanding is that OpenStep is later than NeXTStep. Nextstep was Nexts first OS and includes Mach (which it is riding on) and its own UI. OpenStep is the portable parts that can ride on other OS's (like WinNT). The UI parts of Nextstep are going to be rewritten (or ride on) a more Mac look and feel. The system will still be Mach based. But there will be new adds - like QTML. So it will be a Nextstep derivative with some Mac additions and a new UI. But I am somewhat of a Next-Newbie myself. (I played with a cube for a while in '88)... David K. Every http://MacKiDo.netgate.net/ <- the Power of the Macintosh way From dke@san.rr.com Mon Mar 17 19:41:12 1997 Return-Path: Received: from san.rr.com by omnigroup.com (NX5.67f2/NX3.0M) id AA05176; Mon, 17 Mar 97 19:41:12 -0800 Received: (from uucp@localhost) by mail.san.rr.com (8.7.6/8.7.3) id TAA07086; Mon, 17 Mar 1997 19:41:00 -0800 (PST) Message-Id: <199703180341.TAA07086@mail.san.rr.com> Received: from dt6h2n5a.san.rr.com(204.210.36.90) by mail via smap (V1.3) id tmp007008; Mon Mar 17 19:40:39 1997 Subject: Re: NextStep vs OpenStep Date: Mon, 17 Mar 97 19:51:17 -0800 X-Mailer: Claris Emailer 2.0 x49, February 10, 1997 From: "David K. Every" To: , "Multiple recipients of list" Mime-Version: 1.0 Content-Type: text/plain; charset="US-ASCII" >NextStep is a precursor to OpenStep. > >The former is a NeXT-only technology while the latter has been implemented >by both NeXT and Sun, and is currently being implemented by a collection of >talented folks as part of the GNU initiative. > >OpenStep, being more modern than NextStep, improves on the original in >several areas: Gads... you mean that OpenStep on Mach is not considered Nextstep?!?! I had that wrong. I though OpenStep was Nextstep that was middleware only - and if you had Mach and the Next look and feel it was Nextstep. Appologies for any idiocy I may have stated. David K. Every http://MacKiDo.netgate.net/ <- the Power of the Macintosh way From msb@plexare.com Mon Mar 17 19:43:30 1997 Return-Path: Received: from empire.is.com by omnigroup.com (NX5.67f2/NX3.0M) id AA05270; Mon, 17 Mar 97 19:43:30 -0800 Received: from mspboss.is.com by empire.is.com; Mon, 17 Mar 1997 21:43:16 -0600 Received: by mspboss.is.com with UUCP (8.6.12/IS-Mailhost-1.0.1/plex_net33) id VAA28971; Mon, 17 Mar 1997 21:42:43 -0600 Received: from imladris by plexare.com with SMTP (NX5.67f2/Plexare-1.0.2) id AA04261; Mon, 17 Mar 97 21:47:46 -0600 Message-Id: <9703180347.AA04261@plexare.com> Received: by imladris.plexare.com (NX5.67g/NX3.0X) id AA00655; Mon, 17 Mar 97 21:47:52 -0600 Content-Type: text/plain Mime-Version: 1.0 (NeXT Mail 4.1mach v148) Received: by NeXT.Mailer (1.148) From: "Michael S. Barthelemy" Date: Mon, 17 Mar 97 21:47:51 -0600 To: rhapsody-dev@omnigroup.com Subject: Introductions Reply-To: msb@is.com Since everyone else seems to be introducing themselves I thought I would do the same... I have been a NEXTSTEP developer since January 1989 when I fell in love with a sleek black cube while I was in school. After college, I went out looking for work as a NEXTSTEP developer and finally found it in 1991. Since then I have developed software for NEXTSTEP which covers quite a bit of ground: from two scanner drivers, a driver for film recorder, an image editor, a NEXTSTEP client for HP OpenMail, several toolkits for further easing database application development (one home grown, and two for EOF 1.x), a WOF toolkit, and 24 database applications (yikes - I had never bothered to count!). I have also been a systems administrator for several of the companies I have worked for. I have worked for many of the notable and infamous application and/or consulting firms in the NEXTSTEP: Talus Imaging and Communications (very infamous), Chromagrafx Imaging Systems (Bonus points to those who remember the debates in c.s.n.* about something this company did.), Information Technology Solutions (somewhat infamous), SHL Systemhouse, Integrity Solutions and my current employer Advanced Information Solutions. While at the various consulting firms I have consulted for Swiss Bank Corporation, USWest Communications, and (unknown now but soon to be known) PowerAgent. Plexare is a one person company which I founded in 1993 and I have one shareware application which is out called ClassMaker which assists in generating RTF documentation from source code. (And to those listening who know about the program, I do plan on releasing the 2.0 version but it is more dependant on my finding the time to finish it - it's in that legendary 80-90% done zone and has been there for about two years. Now that there is a larger market perhaps some decent sized customer could request to have it finished! ;-) I enjoy the NEXTSTEP/OPENSTEP development and user environments and was very pleased by Apple's acquisition. I look forward to talking with all of you about all of these exciting developments in our worlds! Mike Barthelemy Senior OPENSTEP Developer/Architect msb@plexare.com From msb@plexare.com Mon Mar 17 19:44:53 1997 Return-Path: Received: from empire.is.com by omnigroup.com (NX5.67f2/NX3.0M) id AA05285; Mon, 17 Mar 97 19:44:53 -0800 Received: from mspboss.is.com by empire.is.com; Mon, 17 Mar 1997 21:43:15 -0600 Received: by mspboss.is.com with UUCP (8.6.12/IS-Mailhost-1.0.1/plex_net33) id VAA28968; Mon, 17 Mar 1997 21:42:43 -0600 Received: from imladris by plexare.com with SMTP (NX5.67f2/Plexare-1.0.2) id AA04226; Mon, 17 Mar 97 21:22:52 -0600 Message-Id: <9703180322.AA04226@plexare.com> Received: by imladris.plexare.com (NX5.67g/NX3.0X) id AA00645; Mon, 17 Mar 97 21:22:58 -0600 Mime-Version: 1.0 (NeXT Mail 4.1mach v148) Content-Type: text/enriched; charset=us-ascii Content-Transfer-Encoding: 7bit Content-Disposition: inline Received: by NeXT.Mailer (1.148) From: "Michael S. Barthelemy" Date: Mon, 17 Mar 97 21:22:57 -0600 To: markm@mail.tyrell.com Subject: Re: Drawing Directly To Screen Cc: Multiple recipients of list Reply-To: msb@is.com References: Under OPENSTEP for Mach and NEXTSTEP before it there were a set of classes known collectively as the InterceptorKit. These allowed you to write directly to the screen without having to go through the DPS server. As I understand, from secondhand information, they did not handle any operations other than writing to the screen. However I imagine someone at Omni can give you a better answer since they had access to the InterceptorKit when they ported Doom ][ to NEXTSTEP. Unfortunately NeXT never released the API's for these kits. I would imagine that they will have to do that now because of greater demand. Someone should start a grass roots movement to start this process - since NeXT has in the past been very stubborn about this issue. So far as regions are concerned this is handled differently by having a full transparency implementation. (Ok, it isn't as comprehensive as what GX has but having written an image editor for NEXTSTEP it was not limiting to only have 13 transparency operators.) The reason it was necessary on the Mac, as I understand it, was because of early memory limitations. Also regions can also be covered by using PostScript user paths. (Check the Adobe Red Book for more details.) I am not certain if you will be able to maintain the portability cross-platform under Rhapsody and OPENSTEP for Mach by using the InterceptorKit. However I imagine that there are a few NeXT engineers which have thought about this issue and probably have a solution in mind. (Although they might not have the opportunity given time constraints.) The InterceptorKit has never been a part of the OPENSTEP specification though, so I doubt that it will be available for OPENSTEP for NT or 95. (It would be nice to be proven wrong though.) Mike Barthelemy Senior OPENSTEP Developer/Architect msb@plexare.com Begin forwarded message: Date: Mon, 17 Mar 97 17:18:54 -0800 Reply-To: markm@mail.tyrell.com Sender: rhapsody-dev@omnigroup.com From: "Mark F. Murphy" < To: Multiple recipients of list < Subject: Drawing Directly To Screen I've been doing game ports for the last several years... so I'm very interested in knowing how to access the screen directly or via a good set of classes. On the Mac, I basically use a PixMap to get info about the port (color table, size, etc). But mostly... a routine called CopyBits is used to copy one PixMap to another. For those unfamilar, CopyBits not only copies from one port to the next... it can reshape... resize... use masks... use transfers modes... dither... transparency, etc. So as you can tell, it does quite a bit. For most things, CopyBits is fine (on a low level)... but at en even lower level, one might write directly into memory described by the PixMap for speed consideration. Also... there's some concepts in QuickDraw on the Mac that come in handy... like Regions (a non-rectangular description of an area). Regions can come in many different shapes and size (donuts, etc). I haven't looked at what QuickDraw GX offered (since it wasn't bundled with the MacOS)... but some say it was the next step up from what Color QuickDraw offered on the Mac. What way will we be able to access graphic devices... and still be able to address multiple monitors and keep cross platform in mind (i.e. the target platform might be Rhapsody or Win NT/95). I'm all ears! mark --------------------------------------------------------------------------- Mark F. Murphy, Director Software Development, Tyrell (markm@tyrell.com) Sysop, The Desktop BBS (Macintosh) 714-491-1003 --------------------------------------------------------------------------- Member of Families Against Internet Censorship: rainbow.rmi.net/~fagin/faic From mpinkert@lennon.cc.gatech.edu Mon Mar 17 19:45:41 1997 Return-Path: Received: from burdell.cc.gatech.edu by omnigroup.com (NX5.67f2/NX3.0M) id AA05293; Mon, 17 Mar 97 19:45:41 -0800 Received: from lennon.cc.gatech.edu (root@lennon.cc.gatech.edu [130.207.9.20]) by burdell.cc.gatech.edu (8.8.4/8.6.9) with ESMTP id WAA20086 for ; Mon, 17 Mar 1997 22:45:33 -0500 (EST) Received: from [130.207.118.117] (fire.cc.gatech.edu [130.207.118.117]) by lennon.cc.gatech.edu (8.8.4/8.6.9) with SMTP id WAA10672 for ; Mon, 17 Mar 1997 22:45:32 -0500 (EST) Message-Id: <199703180345.WAA10672@lennon.cc.gatech.edu> Subject: reply-to field not set? Date: Mon, 17 Mar 97 22:43:45 -0500 X-Sender: mpinkert@lennon.cc.gatech.edu X-Mailer: Claris Emailer 2.0 x52, February 26, 1997 From: Mike Pinkerton To: "rhapsody-dev" Mime-Version: 1.0 Content-Type: text/plain; charset="US-ASCII" I just noticed that the reply-to field in the header is not getting set for any of the messages coming originating from the listserver, so my reply's are going to indiviual people instead of to the list... Is this a problem only I'm having? If not, I think it should be fixed....if this isn't the appropriate place for reporting this kind of error, please let me (and everyone else) know where to do it. Thanks -- Mike Pinkerton mpinkert@cc.gatech.edu http://www.cc.gatech.edu/~mpinkert Cyberdog: On the Internet, no one knows you're an OpenDoc part From lbrooks@apl.bbdo.com Mon Mar 17 20:08:09 1997 Return-Path: Received: from [207.124.223.40] by omnigroup.com (NX5.67f2/NX3.0M) id AA05812; Mon, 17 Mar 97 20:08:09 -0800 Received: from [204.31.15.183] by apl.bbdo.com with SMTP (Apple Internet Mail Server 1.1.1); Mon, 17 Mar 1997 20:15:35 -0800 Subject: NextStep version on cube? Date: Mon, 17 Mar 97 20:04:11 -0800 X-Mailer: Claris Emailer 1.1 From: Leigh Brooks To: Mime-Version: 1.0 Content-Type: text/plain; charset="US-ASCII" Message-Id: <1353492761-203071430@apl.bbdo.com> Hi folks, I have a couple of very basic questions. An '040 cube dropped out of the sky and onto my desk (actually it's been in store room for 5 years - I dug it out and it works. Has OS 2.01.) What version of NextStep can I run and where to get it? (I assume that OpenStep refers to running on Intel hardware). Is what I would be able to run representative of the current state of OpenStep? Thanks, Leigh Brooks From don@misckit.com Mon Mar 17 20:35:02 1997 Return-Path: Received: from misckit.com by omnigroup.com (NX5.67f2/NX3.0M) id AA06011; Mon, 17 Mar 97 20:35:02 -0800 Received: from brain by misckit.com (NX5.67e/NX3.0M) id AA26251; Mon, 17 Mar 97 21:24:46 -0700 Message-Id: <9703180424.AA26251@misckit.com> Received: by brain.misckit.com (NX5.67g/NX3.0X) id AA00824; Mon, 17 Mar 97 21:24:49 -0700 Mime-Version: 1.0 (NeXT Mail 4.1mach v148) Content-Type: text/enriched; charset=us-ascii Content-Transfer-Encoding: 7bit Content-Disposition: inline In-Reply-To: <9703180322.AA04226@plexare.com> X-Nextstep-Mailer: Mail 4.1mach (Enhance 2.0b5) Received: by NeXT.Mailer (1.148.RR) From: Donald A. Yacktman Date: Mon, 17 Mar 97 21:24:48 -0700 To: rhapsody-dev@omnigroup.com Subject: Re: Drawing Directly To Screen Reply-To: don@globalobjects.com References: <9703180322.AA04226@plexare.com> Approved: don@misckit.com X-Warners: Yakko, Wakko, and Dot On Mon, 17 Mar 97, "Michael S. Barthelemy" wrote: > [... Interceptor ...]. Someone should start a grass roots movement to > start this process - since NeXT has in the past been very stubborn about > this issue. Actually, Omni has access because they reverse engineered it, and so they have done the first step of a "grass roots" movement, since could in theory give it away to anyone they want. They may not, simply to keep off NeXT's s*** list, though, so don't pester them too much right now. As another reason not to worry too much about the Interceptor right now, here's something from an Omni guy (posted without his permission, but I seriously doubt he'll mind): On Mon, 17 Mar 97, Greg Titus wrote: > Donald A. Yacktman < says: > > 1. Interceptor is not a panacea. In fact, if you aren't clever > > enough, DPS will outperform your direct-to-screen routines. And it isn't > > as idiot-proof as DPS, so it is easy to get bit--badly. > > I just wanted to underline this statement (as someone who has written code > using Interceptor and DPS). The bitmap display code path in Display > PostScript is very heavily optimized. Interceptor (which goes direct to the > screen) is _maybe_ 10% faster than creating a bitmap object whose image > data matches the bit-depth of the screen and drawing it via DPS. (If DPS > has to dither your image or something it is going to get slower fast, > obviously.) > > There really is little reason to try and get around DPS except in maybe the > most demanding of applications. If you read the docs I listed in my earlier post, you'll find out a lot about those optimizations that DPS makes. I seriously doubt that most Interceptor users will be able to beat them without going into assembly code. It can get really nasty. And, for the (max) 10% you lose going through DPS you gain hardware independence. For 90% of the games out there, IMHO you're better off trading that bit of speed for hardware independence. Only the _most_ demanding stuff needs Interceptor, and you'll have to work REALLY hard to actually get a benefit. Of course, there are applications where it is appropriate--so these warnings are to only make sure developers give it a long hard analysis before they try. As I said before, you can go through DPS and switch in the Interceptor later on, so it is much smarter to go that route. If your game is fast enough on DPS--which is likely--then don't bother doubling your development time to go onto Interceptor. I recommend this also because you are going to _think_ you need Interceptor until you actually go through DPS yourself and discover how fast it really is! > [...Regions map to alpha and DPS user paths...] Right. The MiscKit has a class to ease use of user paths, too. Also, be sure to become intimately familiar with NSImage and NSImageBitmapRep, which can be a bridge between an array of pixels in memory and DPS. By the way, note that you can use DPS with all it's cool type stuff and path stuff to render an offscreen buffer and then use Interceptor or some other method to blit it to the screen. And you can scribble into that same buffer if you want to, just by accessing the data pointer like a C array. In fact, off screen rendering has been used to turn dot matrix printers attached to NeXT OS machines into PS printers! Just render into an offscreen buffer the size of a page and then splat the buffer to the printer. This is also effectively how the black hardware printers operated. > I am not certain if you will be able to maintain the portability > cross-platform under Rhapsody and OPENSTEP for Mach by using the > InterceptorKit. However I imagine that there are a few NeXT engineers which > have thought about this issue and probably have a solution in mind. > (Although they might not have the opportunity given time constraints.) The > InterceptorKit has never been a part of the OPENSTEP specification though, > so I doubt that it will be available for OPENSTEP for NT or 95. (It would > be nice to be proven wrong though.) I think you have to just use those OS's methods for splatting bits. Makes things tough for cross platform usage. I would be interested in seeing where Interceptor goes. If you want true cross platform, though, avoid the Interceptor like the plague. NeXT did the cross platform for you when they build DPS and they've worked hard to optimize it. If you use Interceptor you are effectively saying that you can do a better job than they can. Since their solution is a general one, you probably can do better, but expect it to be a tough benchmark to supercede! --- Later, -Don Yacktman don@misckit.com <My home page< From sanguish@digifix.com Mon Mar 17 21:02:02 1997 Return-Path: Received: from digifix.digifix.com by omnigroup.com (NX5.67f2/NX3.0M) id AA06861; Mon, 17 Mar 97 21:02:02 -0800 Received: (from sanguish@localhost) by digifix.com (8.8.5/8.8.5) id XAA13349 for rhapsody-dev@omnigroup.com; Mon, 17 Mar 1997 23:28:40 -0500 (EST) Message-Id: <199703180428.XAA13349@digifix.com> Content-Type: text/plain Mime-Version: 1.0 (NeXT Mail 4.1mach v148) In-Reply-To: <9703180347.AA04261@plexare.com> X-Nextstep-Mailer: Mail 4.1mach (Enhance 2.0b5) Received: by NeXT.Mailer (1.148) From: Scott Anguish Date: Mon, 17 Mar 97 23:28:39 -0500 To: rhapsody-dev@omnigroup.com Subject: Re: Introductions Reply-To: sanguish@digifix.com References: <9703180347.AA04261@plexare.com> I guess its my turn. I've been developing for NEXTSTEP since I bought my first cube back 6 years ago or so... before that I was a Mac developer since pretty much the start (Aug 84) doing mass market stuff and later vertical products for the textile industry. Right now I'm pretty much involved in WebObjects stuff both during my work hours (I consult for a NeXTSTEP company) and in my other hours (running Stepwise.. more later). I love working under OpenStep, and firmly believe that if we've been able to weather the last few years, this Apple thing should be pretty easy by comparision.. I'm fairly involved in the OpenStep community. I moderate comp.sys.next.announce, I post the weekly 'NEXTSTEP Resources on the Internet' summaries to the groups, and am involved (although to a much less degree that I was) with the ftp.next.peak.org site (formal ORST). On top of all that, I run Stepwise (www.stepwise.com). Stepwise was created as a gathering point for ISV product information when NextWorld magazine stopped publishing. There is information about WWW sites, FTP sites, the newsgroups, OpenStep books, mailing lists, as well as company and product information. Lots there really, and I'm always open to submissions. Since the "merger" I've been keeping an updated set of links to current articles about Rhapsody specifically, and other items that might be of interest. It gets edited pretty much daily. Much of Stepwise is driven by NeXT's WebObjects product, and that percentage is growing every day. Anyway, I'm happy to help any of our new developer friends, or for that matter, any of the current developers in pretty much any way I can. Scott From kc Mon Mar 17 21:03:39 1997 Return-Path: Received: from criamon by omnigroup.com (NX5.67f2/NX3.0M) id AA06966; Mon, 17 Mar 97 21:03:39 -0800 Message-Id: <9703180503.AA06966@omnigroup.com> Received: by criamon.omnigroup.com (NX5.67g/NX3.0X) id AA01276; Mon, 17 Mar 97 21:03:39 -0800 Content-Type: text/plain Mime-Version: 1.0 (NeXT Mail 4.1mach v148) X-Image-Url: ftp://ftp.omnigroup.com/pub/Images/People/kc.tiff X-Nextstep-Mailer: Mail 3.3 (Enhance 1.3) Received: by NeXT.Mailer (1.148) From: Ken Case Date: Mon, 17 Mar 97 21:03:38 -0800 To: mpinkert@cc.gatech.edu Subject: Re: reply-to field not set? Cc: rhapsody-dev Reply-To: kc@omnigroup.com X-Url: http://www.omnigroup.com/People/kc/ X-Organization: Omni Development, Inc. > I just noticed that the reply-to field in the header is not getting > set for any of the messages coming originating from the listserver, > so my reply's are going to individual people instead of to the > list... That's an intentional setting, the default behavior we've always had on the other NeXT mailing lists. However, I don't have any strong feelings one way or the other about the setting, so if people prefer that their replies go to the list by default (rather than to the author of a message), let me know. Ken From msb@plexare.com Mon Mar 17 21:05:13 1997 Return-Path: Received: from empire.is.com by omnigroup.com (NX5.67f2/NX3.0M) id AA06980; Mon, 17 Mar 97 21:05:13 -0800 Received: from mspboss.is.com by empire.is.com; Mon, 17 Mar 1997 23:04:56 -0600 Received: by mspboss.is.com with UUCP (8.6.12/IS-Mailhost-1.0.1/plex_net33) id XAA01888; Mon, 17 Mar 1997 23:04:24 -0600 Received: from imladris by plexare.com with SMTP (NX5.67f2/Plexare-1.0.2) id AA04579; Mon, 17 Mar 97 23:10:28 -0600 Message-Id: <9703180510.AA04579@plexare.com> Received: by imladris.plexare.com (NX5.67g/NX3.0X) id AA00809; Mon, 17 Mar 97 23:10:34 -0600 Content-Type: text/plain Mime-Version: 1.0 (NeXT Mail 4.1mach v148) Received: by NeXT.Mailer (1.148) From: "Michael S. Barthelemy" Date: Mon, 17 Mar 97 23:10:34 -0600 To: don@misckit.com Subject: Re: Drawing Directly To Screen Cc: Multiple recipients of list Reply-To: msb@is.com References: <9703180424.AA26251@misckit.com> > Of course, there are applications where it is appropriate--so these warnings are to only make > sure developers give it a long hard analysis before they try. As I said before, you can go > through DPS and switch in the Interceptor later on, so it is much smarter to go that route. If > your game is fast enough on DPS--which is likely--then don't bother doubling your development > time to go onto Interceptor. I recommend this also because you are going to _think_ you need > Interceptor until you actually go through DPS yourself and discover how fast it really is! I have to agree with this whole-heartedly. My response was meant to be informative -- not persuasive to use the InterceptorKit if/when it becomes available. Mike Barthelemy Senior OPENSTEP Developer/Architect msb@plexare.com From bbum@precipice.com Mon Mar 17 21:15:34 1997 Return-Path: Received: from lux.precipice.com by omnigroup.com (NX5.67f2/NX3.0M) id AA07219; Mon, 17 Mar 97 21:15:34 -0800 Received: (from bbum@localhost) by lux.precipice.com (8.8.5/8.8.5/lux-sharedsub) id AAA07902 for rhapsody-dev@omnigroup.com; Tue, 18 Mar 1997 00:05:50 -0500 (EST) Message-Id: <199703180505.AAA07902@lux.precipice.com> Mime-Version: 1.0 (NeXT Mail 4.1mach v148) Content-Type: text/enriched; charset=us-ascii Content-Transfer-Encoding: 7bit Content-Disposition: inline Received: by NeXT.Mailer (1.148) From: Bill Bumgarner Date: Tue, 18 Mar 97 00:05:49 -0500 To: Multiple recipients of list Subject: Re: Drawing Directly To Screen References: <9703180424.AA26251@misckit.com> X-Truth: The wind is like the air, only pushier. Looking at OpenStep 4.1 for Mach-- in frameworks: NextTime is available as a framework. It is designed to get bitmaps onto screen very quickly-- specifically, it is designed to do just such a thing while also keeping it in sync with audio and/or other events. And it is designed to consume media streams; they can be sourced from files, memory, network or whatever.... ...and the API is available and appears to be documented. Interceptor is also available as a framework-- but its API is not. --- Ultimately, head the advice of the experts... DPS [Display PostScript] really is a very fast, highly optimized imaging system. If you can wrap your head around it and figure out just what form you should convert your image data into, you make it hit the screen in a fast and optimized manner. Because DPS/OpenStep so completely abstracts the display hardware behind a well-defined, consistent, display independent API, going below that API will open up a pandora's box of implementation issues you likely do not want to deal with. For example, the current video card I'm using-- an Elsa Winner Pro w/ 4mb of video-- has approximately 150 different, fully supported video modes... YET-- even with all those modes (and that's just one card!)-- none of my display code has EVER had to deal with that issue... yet, the quality of display is consistent across ALL possible video modes! Imagine-- under OpenStep/Rhapsody, you could be running FrameMaker on a black and white machine and pick pantone colors from the standard, shared system wide, color picker panel and KNOW that: - on a color system, it will display as close as possible to that color given the current output device - upon publishing, it would match exactly - *in any other program* that you happen to paste an EPS snippet of the document OR display the postscript of the document, the color will match. Pretty cool, huh? ...This is going to be fun.... ...I cannot WAIT to play with an OS native version of PhotoShop... b.bum From wjs Mon Mar 17 21:16:14 1997 Return-Path: Received: from foom by omnigroup.com (NX5.67f2/NX3.0M) id AA07226; Mon, 17 Mar 97 21:16:14 -0800 Message-Id: <9703180516.AA07226@omnigroup.com> Received: by foom.omnigroup.com (NX5.67g/NX3.0X) id AA06278; Mon, 17 Mar 97 21:16:13 -0800 Content-Type: text/plain Mime-Version: 1.0 (NeXT Mail 4.1mach v148) In-Reply-To: X-Nextstep-Mailer: Mail 3.3 (Enhance 1.3) Received: by NeXT.Mailer (1.148) From: William Shipley Date: Mon, 17 Mar 97 21:16:11 -0800 To: rhapsody-dev Subject: Re: Drawing Directly To Screen Reply-To: wjs@omnigroup.com References: X-Image-Url: ftp://ftp.omnigroup.com/pub/Images/People/wjs.tiff X-Organization: Omni Development, Inc. OpenStep's rough equivalent of CopyBits would be the postscript level 2 "image" operator, which can copy from one window to another (with scaling, color correction, rotation, whatever). This is optimized for certain types of operations, but it ain't something I've used for games. There is also, in NeXT's DPS, "alphaimage" and "composite", the former can be used to render an image that contains an alpha channel, the latter is a simple and fast way to copy bitmaps between windows, if you aren't interested in changing the rotation or scaling. As for Interceptor, it's currently undocumented, although you can get at the headers with a little tweaking. It's been a couple years since I used it, and the interface has changed a bit since then, but I think it's mostly the same, so I'll outline what I knew and show today's interfaces. In general it looks like you always access a subclass of NSSimpleBitmap, which provides a simple interface to figure out the format of the device you are working with, eg: - (BOOL)hasAlpha; - (BOOL)isPlanar; - (int)bitsPerPixel; - (int)bitsPerSample; - (int)bytesPerPlane; - (int)bytesPerRow; - (int)numberOfPlanes; - (int)pixelsHigh; - (int)pixelsWide; - (int)samplesPerPixel; - colorSpace; and ways to get a handle directly on the data: - (void)getBitmapDataPlanes:(STR *)fp16; - (STR)data; The two subclasses of NSSimpleBitmap are NSDirectBitmap, which represents an area on the screen that you can write directly to, and NSFramebuffer, which represents an entire screen. Using the NSFrameBuffer (eg, if you want to take over the screen), it looks like you create one with: - initFromScreen:(int)screenNumber andMapIfPossible:(BOOL)doMap; Unfortunately, it's a little difficult to get screenNumbers right now, as it's a private variable on the NSScreen class. Assumedly they'll make this method take an NSScreen instance when they make Interceptor public. Some devices allow you to map the screen's video memory in main memory, so that you can read and write directly to the screen. If you pass in TRUE to andMapIfPossible, the NSFrameBuffer you get back will be mapped if the device supports it. If it's not mapped, it'll just allocate some memory somewhere, and when you call -unlock, it'll flush to the real screen memory. Note that whether or not you are mapped doesn't change how you access the NSFrameBuffer. Before you do any writing to the NSFrameBuffer, you have to call - (void)lockWithMode:(int)fp16; I don't know what the valid values for mode are-- I don't think the mode was in the version I used. I imagine there is an enum somewhere that defines modes for reading, writing, and readwrite. When you are done writing (or reading), you must call - (void)unlock; which will flush your changes to the screen if this is an unmapped framebuffer, and will in any event let PostScript start writing to the screen again. In the old days, PostScript could NOT write to the screen while you had it locked, so you didn't want to leave these babies locked forever (or you'd starve out the windowserver). I assume this is still true. You can determine how to write to the framebuffer by asking it for its - (NSString *)pixelEncoding; Which returns a string describing the format of the framebuffer's memory, like "RRRRGGGGBBBB----" for 4 bits red, 4 bits green, 4 bits blue, and 4 bits unused. It's up to you to write drawing routines that can handle every permutation of pixel encodings in the world. This is a big drawback. Another big drawback is PostScript can't write to NSFrameBuffers directly, so if your framebuffer isn't mapped, you can't mix PostScript and framebuffer access (because when you flush your framebuffer, you'll overwrite the changes you made to the real screen using PostScript). There really ought to be some way to take the mapped memory and create a PostScript window out of it, so you can access it with PostScript and with your own graphics calls directly. If you aren't taking over the full screen, you'll use the NSDirectBitmap subclass, which is initialized with - initForRect:(NSRect)rect inWindow:(NSWindow *)window; Eg, you just specify a region in some window you want direct access to. This is pretty darn convenient. You can call - (void)setDirectMapped:(BOOL)directMapped; to make the NSDirectBitmap actually represent the memory on screen or not -- I don't know what the default is. Again, if you say not, it'll give you an offscreen buffer that gets flushed when you call -flush (and possible -unlockBitmap, see below). There's also a - (void)setBuffered:(BOOL)buffered; And I honestly don't know how buffered mode differs from directMapped mode (what if you are buffered and not directmapped? Are you buffered twice? Who knows? Well, the NEXTIME guys do). Again, before you write to your bitmap you want to call - (void)lockBitmap; And then when you are done - (void)unlockBitmap; There's also - (void)flushIn:(NSRect)rect; - (void)flush; Which I think only mean anything if you are in buffered mode. Again, to write to your NSDirectBitmap you want to ask it for its -pixelEncoding, and then for a pointer to its -data, and then you can just start blatting away, assuming you have routines to write to that pixel encoding. When we ported DOOM II to Interceptor, Carmack had represented all pixels as 8-bit offsets into a 24 bit color lookup table, and so we wrote a bunch of little routines that would look up the 24-bit RGB values up in a dither table for each kind of depth we supported. Back then there were only about 5 different types of framebuffers to support (2-bit gray, RGB:444, RGB:888, and BRG:888), so it wasn't that hard. Nowadays there are a ton more devices supported, so you could be in a big stinky world of hurt. The biggest limitations of Interceptor right now are: 1) Not supported by NeXT. On the other hand, it's the basis for NEXTIME, so it ain't going away (although the API may change 100%). 2) Can't mix PostScript and Interceptor in the same on-screen rectangle 100% of the time, as far as I know. 3) You have to write low-level routines to write to every different kind of framebuffer in the world. Sigh. 4) You can't dynamically change the screen resolution, so you're stuck with whatever depth and size the user selected. Ouch. -Wil From kc Mon Mar 17 21:19:51 1997 Return-Path: Received: from criamon by omnigroup.com (NX5.67f2/NX3.0M) id AA07391; Mon, 17 Mar 97 21:19:51 -0800 Message-Id: <9703180519.AA07391@omnigroup.com> Received: by criamon.omnigroup.com (NX5.67g/NX3.0X) id AA01336; Mon, 17 Mar 97 21:19:28 -0800 Content-Type: text/plain Mime-Version: 1.0 (NeXT Mail 4.1mach v148) X-Image-Url: ftp://ftp.omnigroup.com/pub/Images/People/kc.tiff X-Nextstep-Mailer: Mail 3.3 (Enhance 1.3) Received: by NeXT.Mailer (1.148) From: Ken Case Date: Mon, 17 Mar 97 21:19:27 -0800 To: lbrooks@apl.bbdo.com Subject: Re: NextStep version on cube? Cc: rhapsody-dev Reply-To: kc@omnigroup.com X-Url: http://www.omnigroup.com/People/kc/ X-Organization: Omni Development, Inc. > An '040 cube dropped out of the sky and onto my desk (actually it's > been in store room for 5 years - I dug it out and it works. Has OS > 2.01.) What version of NextStep can I run You can run the very latest release: OpenStep/Mach 4.1. (And presumably you'll also be able to run 4.2 when it comes out.) > and where to get it? http://www.optimal-object.com/ Ken From robertl@esplanade.hot.com Mon Mar 17 21:38:28 1997 Return-Path: Received: from toto.pn.com by omnigroup.com (NX5.67f2/NX3.0M) id AA07587; Mon, 17 Mar 97 21:38:28 -0800 Received: from pioneer.ci.net (pioneer.ci.net [205.136.67.65]) by toto.pn.com (8.7.5) with ESMTP id AAA05010 for ; Tue, 18 Mar 1997 00:38:18 -0500 (EST) Received: (from Uesplanade@localhost) by pioneer.ci.net (8.8.2/8.8.0) with UUCP id AAA13337 for rhapsody-dev@omnigroup.com; Tue, 18 Mar 1997 00:37:54 -0500 X-Authentication-Warning: pioneer.ci.net: Uesplanade set sender to robertl@esplanade.hot.com using -f Received: (from robertl@localhost) by esplanade.hot.com (8.8.5/8.8.5) id AAA07905 for rhapsody-dev@omnigroup.com; Tue, 18 Mar 1997 00:32:59 -0500 (EST) From: Robert La Ferla Message-Id: <199703180532.AAA07905@esplanade.hot.com> Mime-Version: 1.0 (NeXT Mail 3.3 v118.2) Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: quoted-printable Received: by NeXT.Mailer (1.118.2) Date: Tue, 18 Mar 97 00:32:58 -0500 To: rhapsody-dev@omnigroup.com Subject: My Introduction... Reply-To: Robert La Ferla References: <9703180347.AA04261@plexare.com> Since everyone seems to be forwarding their NeXT history, I will too. I've = been working with NEXTSTEP for over seven years (since version 0.9) and MACH = OS for a couple more. After working at BBN on MACH OS development, I went = on to Lotus Development to work in the groundbreaking Lotus Improv = spreadsheet group. I have since been developing ObjectWare for NEXTSTEP = namely the BarCodeKit (NeXTWorld Best of Breed nominee) and SerialPortKit. = But my focus in the past few years has been consulting. I've worked on = NEXTSTEP front-ends for real-time embedded systems to control semiconductor = equipment (used to manufacture memory chips and Pentium CPUs) for Lam = Research and most recently with Fannie Mae, the nation's largest diversified = financial company developing mortgage applications for the Object Factory. = I've taught NeXT developer courses for NeXT and have provided OOA/OOD/OOP = process and technical mentoring services for the Object Systems Group. I = also have extensive system and network administration experience. Many = sites have used my services or my ports of sendmail 8.x, kermit, flex/bison, = and gawk. Lastly, I am the proud owner of a NeXT '040 cube w/floppy (rare), = a kick-butt 200MHz Pentium Pro system, a Pentium, and three Apple Newtons = (MP120, MP130 and the new MP2000.) I also have an Atari 800 and Amiga = somewhere in my attic... I hope to purchase a SMP PowerPC system in the = future!!! I am very excited about working with Mac developers, Rhapsody and Apple = hardware. I am anxious to learn as much about the Mac as you Mac developers = are anxious to learn about OPENSTEP. Feel free to drop me a note. Robert La Ferla Registered OPENSTEP Developer/Consultant NeXT and Newton aficionado Robert_La_Ferla@hot.com From erbenson@akcache.com Mon Mar 17 22:04:06 1997 Return-Path: Received: from MAILHOST.AKCACHE.COM by omnigroup.com (NX5.67f2/NX3.0M) id AA07917; Mon, 17 Mar 97 22:04:06 -0800 Received: from [206.66.199.114] by mailhost.akcache.com (Netscape Mail Server v1.1) with SMTP id AAA158 for ; Mon, 17 Mar 1997 21:07:17 -0900 Subject: PPC 601 and NuBus based systems Date: Mon, 17 Mar 97 21:04:08 -0900 X-Sender: erbenson@mailhost.akcache.com X-Mailer: Claris Emailer 1.1 From: Ethan Benson To: "Rhapsody dev Omnigroup" Mime-Version: 1.0 Content-Type: text/plain; charset="US-ASCII" Message-Id: <19970318060715051.AAA158@[206.66.199.114]> Hello, I hope this is not too off topic... but can the NeXT people give any info about how OpenStep/Rhap can support PPC 601 and Nubus PPC machines? is it possible is it very likly that apple will not complete support for thses machines? no FUD please thanks Ethan Benson Macintosh System Exorcist ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Encrypted mail preferred To get PGP key, send a message with subject "Send FileCrypt Key" PGP Key also available on Keyservers Key FingerPrint: 64 B9 D0 A6 4A 73 87 73 C1 DE 4A 2C 54 F6 0F E1 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ From erbenson@akcache.com Mon Mar 17 22:19:01 1997 Return-Path: Received: from MAILHOST.AKCACHE.COM by omnigroup.com (NX5.67f2/NX3.0M) id AA08120; Mon, 17 Mar 97 22:19:01 -0800 Received: from [206.66.199.114] by mailhost.akcache.com (Netscape Mail Server v1.1) with SMTP id AAA133; Mon, 17 Mar 1997 21:22:24 -0900 Subject: Re: PPC 601 and NuBus based systems Date: Mon, 17 Mar 97 21:19:16 -0900 X-Sender: erbenson@mailhost.akcache.com X-Mailer: Claris Emailer 1.1 From: Ethan Benson To: "Garance A Drosehn" Cc: "Multiple recipients of list" Mime-Version: 1.0 Content-Type: text/plain; charset="US-ASCII" Message-Id: <19970318062223798.AAA133@[206.66.199.114]> >Only Apple can answer that, not a mailing list of 3rd party >developers. If we say it's possible, and Apple does not do >it, then who cares what we've said? This ball is entirely >in Apple's court. I understand this I was only asking if there are any technical issues in OpenStep that would casue problems, I was not asking for a apple prediction. sorry... Ethan Benson Macintosh System Exorcist ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Encrypted mail preferred To get PGP key, send a message with subject "Send FileCrypt Key" PGP Key also available on Keyservers Key FingerPrint: 64 B9 D0 A6 4A 73 87 73 C1 DE 4A 2C 54 F6 0F E1 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ From gad@eclipse.its.rpi.edu Mon Mar 17 22:34:57 1997 Return-Path: Received: from mail1.its.rpi.edu by omnigroup.com (NX5.67f2/NX3.0M) id AA08288; Mon, 17 Mar 97 22:34:57 -0800 Received: from eclipse.its.rpi.edu (gad@eclipse.its.rpi.edu [128.113.24.33]) by mail1.its.rpi.edu (8.8.5/8.8.5) with SMTP id BAA30066; Tue, 18 Mar 1997 01:13:31 -0500 Received: by eclipse.its.rpi.edu (NX5.67e/NX3.0M) id AA14627; Tue, 18 Mar 97 01:13:29 -0500 Message-Id: <9703180613.AA14627@eclipse.its.rpi.edu> Mime-Version: 1.0 (NeXT Mail 3.3 v118.2) Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Received: by NeXT.Mailer (1.118.2) From: Garance A Drosehn Date: Tue, 18 Mar 97 01:13:27 -0500 To: erbenson@MAILHOST.AKCACHE.COM Subject: Re: PPC 601 and NuBus based systems Cc: Multiple recipients of list References: <19970318060715051.AAA158@[206.66.199.114]> > I hope this is not too off topic... > > but can the NeXT people give any info about how OpenStep/Rhap > can support PPC 601 and Nubus PPC machines? is it possible is > it very likly that apple will not complete support for these > machines? Only Apple can answer that, not a mailing list of 3rd party developers. If we say it's possible, and Apple does not do it, then who cares what we've said? This ball is entirely in Apple's court. --- Garance Alistair Drosehn = gad@eclipse.its.rpi.edu Senior Systems Programmer (MIME & NeXTmail capable) Rensselaer Polytechnic Institute; Troy NY USA From wjs Mon Mar 17 23:34:54 1997 Return-Path: Received: from foom by omnigroup.com (NX5.67f2/NX3.0M) id AA09020; Mon, 17 Mar 97 23:34:54 -0800 Message-Id: <9703180734.AA09020@omnigroup.com> Received: by foom.omnigroup.com (NX5.67g/NX3.0X) id AA06458; Mon, 17 Mar 97 23:34:53 -0800 Mime-Version: 1.0 (NeXT Mail 4.1mach v148) Content-Type: text/enriched; charset=us-ascii Content-Transfer-Encoding: 7bit Content-Disposition: inline In-Reply-To: <199703180341.TAA07086@mail.san.rr.com> X-Nextstep-Mailer: Mail 3.3 (Enhance 1.3) Received: by NeXT.Mailer (1.148) From: William Shipley Date: Mon, 17 Mar 97 23:34:52 -0800 To: dke@san.rr.com Subject: Re: NextStep vs OpenStep Cc: Multiple recipients of list Reply-To: wjs@omnigroup.com References: <199703180341.TAA07086@mail.san.rr.com> X-Image-Url: ftp://ftp.omnigroup.com/pub/Images/People/wjs.tiff X-Organization: Omni Development, Inc. NEXTSTEP generally refers to NEXTSTEP 3.3, which is pre-OpenStep. However, occasionally people use it to refer to NEXTSTEP 4.0, which is OpenStep compliant. OpenStep refers to an open API standard. "OpenStep for Mach" and "NEXTSTEP 4.0" are the exact same thing -- the latest version of NEXTSTEP with the Mach kernel under it. The former name is preferred, but since NeXT just renamed it, most people haven't changed their speaking over yet. Thus, as far as I know, NeXT doesn't sell a product labelled NEXTSTEP anymore. Clear as mud? -Wil From wjs Mon Mar 17 23:47:58 1997 Return-Path: Received: from foom by omnigroup.com (NX5.67f2/NX3.0M) id AA09254; Mon, 17 Mar 97 23:47:58 -0800 Message-Id: <9703180747.AA09254@omnigroup.com> Received: by foom.omnigroup.com (NX5.67g/NX3.0X) id AA06482; Mon, 17 Mar 97 23:47:57 -0800 Content-Type: text/plain Mime-Version: 1.0 (NeXT Mail 4.1mach v148) In-Reply-To: <9703180613.AA14627@eclipse.its.rpi.edu> X-Nextstep-Mailer: Mail 3.3 (Enhance 1.3) Received: by NeXT.Mailer (1.148) From: William Shipley Date: Mon, 17 Mar 97 23:47:56 -0800 To: Multiple recipients of list Subject: Re: PPC 601 and NuBus based systems Reply-To: wjs@omnigroup.com References: <9703180613.AA14627@eclipse.its.rpi.edu> X-Image-Url: ftp://ftp.omnigroup.com/pub/Images/People/wjs.tiff X-Organization: Omni Development, Inc. Ethan Benson writes: > but can the NeXT people give any info about how OpenStep/Rhap > can support PPC 601 and Nubus PPC machines? Garance A Drosehn writes: > Only Apple can answer that, not a mailing list of 3rd party > developers. On the other hand, it's worth mentioning that OpenStep/Mach 4.1 already DOES support the Nubus, so I can't see any technical reason why Rhapsody couldn't. However, there is no doubt it would take some amount of work to support the Macs particular implementation (and to support it using DriverKit, since the old Nubus drivers aren't part of DriverKit as far as I know), and I'm sure right now the porting team is trying to commit to as little work as possible, so they can ship on time. After they ship, I damn well would expect they'll port Rhapsody more broadly, to every 68040 and better Mac. If not, well, I'll be pouty. -Wil From fischer@fokus.gmd.de Mon Mar 17 23:48:50 1997 Return-Path: Received: from mailhub.fokus.gmd.de by omnigroup.com (NX5.67f2/NX3.0M) id AA09263; Mon, 17 Mar 97 23:48:50 -0800 Received: from unkerich.fokus.gmd.de (unkerich [192.35.149.175]) by mailhub.fokus.gmd.de (8.8.5/8.8.5) with SMTP id IAA16607 for ; Tue, 18 Mar 1997 08:48:23 +0100 (MET) Received: by unkerich.fokus.gmd.de (NX5.67f2/NX3.0S) id AA03447; Tue, 18 Mar 97 08:47:13 +0100 Date: Tue, 18 Mar 97 08:47:13 +0100 From: Robert Fischer Message-Id: <9703180747.AA03447@unkerich.fokus.gmd.de> X-Nextstep-Mailer: MediaMail V4.5.4 To: rhapsody-dev@omnigroup.com Subject: Re: Moving to Rhapsody from Smalltalk For me the significant difference between ObjC and Smalltalk is the speed of development. In Smalltalk I have NO compile time. While debugging a running program I change variables or even code and 'snap' it works on. Of course: that is good for 'hackers' like me ;-) I remember the term 'incremental compiler' (or was it a incr. linker?) which should have made part of OPENSTEP. What's about this? I was a NeXTSTEP programmer. I had to switch to Smalltalk 1.5 years ago and now: I love it! The environment is a little bit 'out-of-date' (VisualWorks) but all I hope is to get Smalltalk for OPENSTEP/Rhapsody or whatever its name will be. Robert ----- - .-. -- -- --- / \ ---- Robert Fischer .-. / \ --- .-. __o .-. @ / \ / \ / \ _`\<,_ / \ GMD-Fokus / \ / \ / \ (*)/ (*) / `-------------- / `---' `-' `-----------' From mhoover@minerva.athenet.net Mon Mar 17 23:48:57 1997 Return-Path: Received: from ns3.athenet.net by omnigroup.com (NX5.67f2/NX3.0M) id AA09276; Mon, 17 Mar 97 23:48:57 -0800 Received: from [143.44.248.22] (hooverm.student.lawrence.edu [143.44.248.22]) by athenet.net (8.7.5/8.7.5) with SMTP id BAA29607 for ; Tue, 18 Mar 1997 01:48:33 -0600 (CST) Message-Id: <199703180748.BAA29607@athenet.net> Subject: Re: Introductions Date: Tue, 18 Mar 97 01:54:27 -0600 X-Sender: mhoover@pop.athenet.net X-Mailer: Claris Emailer 1.1 From: Michael Hoover To: "Multiple recipients of list" Mime-Version: 1.0 Content-Type: text/plain; charset="US-ASCII" Well, um, I guess the obligatory introduction is in order: I am not a NeXT developer. I have never even used a NeXT machine (cube or otherwise). I have not read the OpenStep documentation. I have done a fair amount of Mac development. I do not develop Mac software for a living. I am a senior in college who will be working for Motorola come July. I have made it a daily part of my life to try to keep up with everything Macintosh. When I miss a day it takes two days to get caught up. I unluckily missed this last Friday. I think I'm caught up now :-) I am currently in the process of collecting as much knowledge as I can about software development. It seemed natural that I should subscribe to this list. Those 'other' lists seem rather useless at the moment. I have one question about OpenStep: How is music information handled (MIDI)? Are there music apps available? I would imagine Opcode would port OMS (unfortunately). That is about it. I am basically here to lurk and learn. Michael Hoover From don@misckit.com Tue Mar 18 03:15:57 1997 Return-Path: Received: from misckit.com by omnigroup.com (NX5.67f2/NX3.0M) id AA11656; Tue, 18 Mar 97 03:15:57 -0800 Received: from brain by misckit.com (NX5.67e/NX3.0M) id AA29208; Tue, 18 Mar 97 04:12:32 -0700 Message-Id: <9703181112.AA29208@misckit.com> Received: by brain.misckit.com (NX5.67g/NX3.0X) id AA00294; Tue, 18 Mar 97 04:12:31 -0700 Content-Type: application/x-nextmail Mime-Version: 1.0 (NeXT Mail 4.1mach v148) Next-Attachment: .tar.92.MIDI__was_Re__Introd.attach, 0, 1/1, 0, 0 X-Nextstep-Mailer: Mail 4.1mach (Enhance 2.0b5) Received: by NeXT.Mailer (1.148.RR) From: Donald A. Yacktman Date: Tue, 18 Mar 97 04:12:29 -0700 To: rhapsody-dev@omnigroup.com Subject: MIDI (was Re: Introductions) Reply-To: don@globalobjects.com Approved: don@misckit.com X-Warners: Yakko, Wakko, and Dot begin 600 .tar.92.MIDI__was_Re__Introd.attach M'YV0+EX`&$BPH,&#"!,J7,BPH4.$(&C`N%&C!@@`(#*"D''CHL:,,#Q^'!D2 MA(T9,6;0N$%#Q@R+&6_$J#$#8XV'.'/JW,D30)TY=,+(R0B`S!LW/17>05.F M#)ND4*-*G4JUJM6':=R0*8/'A1PZ9JZ*'0L@!@P8-FC0$)F1(UN0;S7>H!@C MX\F4*UN^S&B6+D889`,+]@E4*%&C2*TN;?ITL./'D"-+5?#$#8@F1UF`B-&Q MB>$<-S0W23,&31BG()"\>6.GS-`[J4`9KF M*(@P8M[4H0/B"9PR;J;,AJ,C]9L[(-+,`='F)^GL;LR\D=,F#)WBEDUK95.& M#`C;(%`T24(D28H?"A0XD7V<#1L0=)C&W!U.L>'"@6(L!R!3%" M1E0PM:%0OP7XVQI-/2?'=F^8L>!O37@WQA)I6`@B'`W2<2((9`CE1GL@N$': M;V(TV9N1=XRWQHT@"*&@EP^&0649K;'QAI``KL;&BUJ!((=Y#/8F!AT__0:C M&2("128*3^S9)VY-(/'$#KVY,<9O>2@'PAEE,->G>W2\D9V3/C9XPG9AC,%G M&/[M"(*;S&5J9QECO-%&&]"YU^2!+J2041+,&34<"$1,`06$>;CAXQQIUID5 M4'+4`:NP(+3J!K)UB(K><=L1Z!^UE]&7!)WNA2G'&D1:!L40O0WE1!E84*&; M4&3(/RB'U M'%CB2^F5=_[7;VYFE&%>'0W.G-_&('2I,Y1OB,'>AP\&"..YZ7YZ6:A./^@& M?VHV2&H+Y\':W!2:S:$IOH,#F$8;6>G\X!=7"OS%JJ1.+7D9-,\KJ=K,D?J? M;'F&,2&H=:H1ZL<2[MRZ\7!H#F5Q`)^H&;T`Y@$':<)O2"ES@%X)@I#BD3?C MH]NQ"5VC(-J``PP2:98@<^I;EK!NI(TA(:@-0GC>M12:`[%V-RA<:,KC*-[EYS9!Z%^G?M,= M8HWA.,<[CAW"\#5`XD!@%'M=8!BCW(PU0+01:C,@0'.H]*Y81>!"%?"J>0-D^5B+.,E,(*H M5:]RC=2H!KDSW*D-M,+A%(P)S-^89CNJ,YWHO*FE1W'(<'Q*7.JBY305[@PY M7VO2AE0U!M'-!@3T&LIU(,:&.IRA!5D1XD5O.9LA#2H+DH(AE305)JVXQ@QJ M#%'.JM6T!;V!;_E<'%/RYL:W-A!0E0\2 M$L9.QLB;@B\,XOP)"7>*JN/42:4.LZE'T<"<:SI43:TK@QD*)ZHPO@$/+0#I M@+[V'Y)>#:W:\I_H/C0C3"&.G<\@8R;`A[ZI$5'3;)H:^5=#M>@F=0:XJ8$^1*4VI"E?P4 MI*:*(G2/];PGK(;2*C6ZAV)L^UJW:CJK6P6/#:I;%Z@.BAU24:QZ:(#=(Y%E MN`_]S37F&<\D-R$W9A:RJ(?DU!6(5PA)K7(.PU9!6G$"!4!K.H"5` MP;"LXGO0&%1W5/?TM;+%.DYY]*#!.01K6)63$ZWFURB[I<&[*;0,;(R'4>4, MA2EA\#`1S4:SP&YG*^%3FMU@-2/P,18Z9\A*4X84U\\%[[AU92R^S$!)9ET3 MQU^A6,2,Z:`$.Z$XS!>`X@>!0 M+ZYQO=S%+%5EI37,4M.5L&-B83&%6-N!%=QLE)_%#L7)YD%/%3FIN2$(KS@9 M@QZ9I@#4K2[-E*A][:[.>0::7HMO#-3KC"C573WO@#>7$FFDC:,Z,]#!7?U# M;!D6TZ#9I@J$J$4>>&Q+AC3L;GRQL2=6:0;>[#V';V_LZC616*X)9X5J\`Q@ MD]!#TZ%ZZ<-SX]*!&TN'!L'*;AB#PX:$ZC1\!2=W4[HFOA*G1CHH%016<,V& M(%*F!%:V[3]FU1>0HX,T<+2PJ5` M*B--(5IKN(Y7W=,$SOK,UKT1V`:/X!I">M"XO*RMJW8[S_^H$MS-=G+CXA98 M)R4W-S!T+>Z\F63>S&A#&B^KD=A$JCK,VC+7;%)<69Q:KAD%6(8>SH.2!B00 M#('#==)5+GDIL'W'40XPK6O6E@*CW&"93$-P35`RM.X*.I]R@WP"E#FM=Z7#<('9854K/CVX*&12G@)-+I#ZY)`'4ML9 M:H;[6)$VV3\"]4]-1;V35EP%23C`S)I&PA<4`%;(3%G&UG`HIAQ>YFOI3Q;?M7>Q!QKB64+D-^12W27.5V2K$UPDY\Y;3X(]WTLD'UT7PM M@'TN4!@T,B(NT![U8GMND'*'Y'UIE$>O$P9))C[.T2I;P39X1"(5\AZW\01- M<$$H<%3EPR=N8!Y.D0]9U@Z(TZ*I!YLTTF\03%GXUN-(VY2 ML1"31MU-C MR!U"`22C95\*$%,ROP]`*(X0)DU09L(`(^T`0; M8D6F4R*4P@,O$`8^D!^2\8O`&(S".(S$6(S&>(S(F(S*N(S,V(S.^(S0&(W2 M.(W46(W6>(W8F(W:N(W^(W@&([B.([D6([F>([HF([JN([LV([N^([P =&(_R.(_T6(_V>(_XF(_ZN(_\V(_^^(\`&9`"^1CF ` end From don@misckit.com Tue Mar 18 03:27:30 1997 Return-Path: Received: from misckit.com by omnigroup.com (NX5.67f2/NX3.0M) id AA11849; Tue, 18 Mar 97 03:27:30 -0800 Received: from brain by misckit.com (NX5.67e/NX3.0M) id AA29320; Tue, 18 Mar 97 04:24:05 -0700 Message-Id: <9703181124.AA29320@misckit.com> Received: by brain.misckit.com (NX5.67g/NX3.0X) id AA00314; Tue, 18 Mar 97 04:24:05 -0700 Content-Type: text/plain Mime-Version: 1.0 (NeXT Mail 4.1mach v148) X-Nextstep-Mailer: Mail 4.1mach (Enhance 2.0b5) Received: by NeXT.Mailer (1.148.RR) From: Donald A. Yacktman Date: Tue, 18 Mar 97 04:24:04 -0700 To: rhapsody-dev@omnigroup.com Subject: MIDI (was Re: Introductions) Reply-To: don@globalobjects.com Approved: don@misckit.com X-Warners: Yakko, Wakko, and Dot Sorry about the NeXTMail message. Here's what I *meant* to send: On Mon, 17 Mar 97, Michael Hoover wrote: > I have one question about OpenStep: How is music information handled > (MIDI)? Not all that well...but there is support, mainly from a third party, CCRMA, at Stanford. They are the keepers of the MusicKit, a pretty darned nice bit of work. But it is a developer tool, and rather obtuse at first. (Obtuse IMHO; once you get used to it, there's actually a lot to recommend it...) It does DSP synthesis and instrument construction as well as MIDI, and works on PC or NeXT hardware. If you have the MusicKit MIDI driver installed, then you'll be in pretty good shape. The trick is to have cards that have MusicKit drivers on PCs. I'm using the Ensoniq Soundscape Elite with a special driver from Brian Willoughby and it works quite well...though I've only excercised the general MIDI features. The big problem is that NeXT's Mach is not a real-time OS, so the MIDI timing is _never_ realiable. If you quit all other apps and jack up the MIDI app's priority, you typically get fine performances even on a 68040, but even then hiccups are still possible. (It's that "preemptive" part of the OS getting in the way. This is one of those rare cases where that can cause problems.) I've had little or no troubles on my Pentium Pro 200 system, though. And even the 68040 was bearable, so I would expect a PowerPC system should give acceptable performance. > Are there music apps available? Yes. There's a simple MIDIPlayer.app which plays back general MIDI files and there is Sequence.app, a sequencer. (Now free; I think I'm one of the few that actually paid money for the thing back when it was a commercial program...) Sequence has some really nice features, such as the ability to create your own plug-in note filters. (You can do wonderful things with those...) The MusicKit comes with a few fun apps. Ensemble.app is perhaps the most famous of the lot and can be thought of as a MIDI effects box--it will filter a MIDI stream and turn it into something else. For example you can have it harmonize a melody you hand it. The filters it comes with don't do a lot, but a creative programmer could go wild with it. It also has a few algorithmic stream generators, so if you've never listened to a fractal melody, here's your chance... The UNIX music tools have, mostly, been ported over, too. The most significant one is csound and it is an amazing synthesis tool...but only if you can wrap your head around it. It is definitely meant for engineers--not at all for the faint of heart. I have found it fun to play with, though. :-) This is definitely the tool to use if you want to invent a new synthesis method. For notation, there's Calliope.app. Some of the ways it does things will seem strange if you;ve used notation software elsewhere, but this is an incredible program. I've typeset a lot of my original compositions with it and I've been extremely happy with the quality of the results. Very professional. It is freeware, written by William Clocksin, a music professor in England (Cambridge I think). Finally, a suite of music apps is produced by Ralf Suckow and Melonsoft over in Germany. It is a commercial app and I've not bought it, so I can't really give any sort of evaluation of it--but it can do synthesis like Csound does and is very powerful from what I hear. Certainly worth looking into, especially if the free stuff doesn't meet your needs. I think that is everything to speak of. There were a few random apps on one of the Princeton ftp servers, but I don't know if that server is still around. CCRMA's is the goldmine for music lovers. (ccrma-ftp.stanford.edu I think) > I would imagine Opcode would port > OMS (unfortunately). Well, anything more than we've got would be better than the little we have right now. :-) Actually, when you consider it, there's a lot more music stuff for NeXTs than should be expected given the demographics of the NeXT market... --- Later, -Don Yacktman don@misckit.com My home page From onevision.de!guenther@genias.de Tue Mar 18 05:04:40 1997 Return-Path: Received: from GIMLI.genias.de by omnigroup.com (NX5.67f2/NX3.0M) id AA12578; Tue, 18 Mar 97 05:04:40 -0800 Received: from onevision.de by gimli.genias.de with uucp (Smail3.1.28.1 #30) id m0w6yaB-00000LC; Tue, 18 Mar 97 14:05 MET Received: from renoir by onevision.de (NX5.67f2/NX3.0M) id AA28260; Tue, 18 Mar 97 13:47:32 +0100 Message-Id: <9703181247.AA28260@onevision.de> Received: by renoir (NX5.67g/NX3.0X) id AA04798; Tue, 18 Mar 97 13:47:33 +0100 Mime-Version: 1.0 (NeXT Mail 4.1mach v148) Content-Type: text/enriched; charset=us-ascii Content-Transfer-Encoding: 7bit Content-Disposition: inline Received: by NeXT.Mailer (1.148) From: Guenther Fuerthaller Date: Tue, 18 Mar 97 13:47:32 +0100 To: rhapsody-dev@omnigroup.com Subject: Re: Introductions Reply-To: guenther@onevision.de References: <199703180748.BAA29607@athenet.net> Well, after anybodys introduction... I work with NextStep since 1991. We, that is OneVision GmbH, are developing one of the last actively supported DTP applications for NextStep named OneVision. This originally has been designed to be a totally integrated type setting as well as image processing, line art and layout tool. Today we are concentrating on a functionality called DigiScript, which allows to import any PS, EPS or PDF document for last minute changes/corrections. In particular this DigiScript function is my child and thus they call me their PostScript guru here. NextStep people might know me better as author of PDFView, a free PDF reader for NextStep. Maybe i am one of few who can represent a kind of software development which many Mac developers are very interested in. Though i am a very lazy poster i am open to and interested in your questions/discussions. Guenther From gbayley@zonk.geko.net.au Tue Mar 18 05:14:35 1997 Return-Path: Received: from netra.geko.net.au by omnigroup.com (NX5.67f2/NX3.0M) id AA12719; Tue, 18 Mar 97 05:14:35 -0800 Received: from [203.2.239.203] (dialup203.geko.net.au [203.2.239.203]) by netra.geko.net.au (8.6.9.Rob+p1/8.6.9) with SMTP id XAA18715 for ; Tue, 18 Mar 1997 23:15:06 +1000 Message-Id: <199703181315.XAA18715@netra.geko.net.au> Subject: Digest version available? Date: Wed, 19 Mar 97 00:14:28 +1100 X-Sender: gbayley@zonk.geko.net.au X-Mailer: Claris Emailer 2.0, March 15, 1997 From: Grant Bayley To: "Multiple recipients of list" Mime-Version: 1.0 Content-Type: text/plain; charset="US-ASCII" Hi all, Even though Emailer 2.0 is doing a great job of filtering all the messages into the one folder, I've gotta ask the question - is there a digest version available? Thanks in advance for any info, Grant Bayley __________________________________________________ Grant Bayley AusMac Archive Admin, Mac Evangelist Spare Time Mac Internet Consultant http://mac.unsw.edu.au/ gbayley@geko.net.au __________________________________________________ From gbayley@zonk.geko.net.au Tue Mar 18 05:14:37 1997 Return-Path: Received: from netra.geko.net.au by omnigroup.com (NX5.67f2/NX3.0M) id AA12725; Tue, 18 Mar 97 05:14:37 -0800 Received: from [203.2.239.203] (dialup203.geko.net.au [203.2.239.203]) by netra.geko.net.au (8.6.9.Rob+p1/8.6.9) with SMTP id XAA18718 for ; Tue, 18 Mar 1997 23:15:10 +1000 Message-Id: <199703181315.XAA18718@netra.geko.net.au> Subject: Lurk and Learn Date: Wed, 19 Mar 97 00:14:31 +1100 X-Sender: gbayley@zonk.geko.net.au X-Mailer: Claris Emailer 2.0, March 15, 1997 From: Grant Bayley To: "Multiple recipients of list" Mime-Version: 1.0 Content-Type: text/plain; charset="US-ASCII" Michael Hoover altered the fabric of cyberspace by writing: >lurk and learn Heheh.. The two words I've lived by in Semper.Fi and now in this list, atleast until I finish Uni and get some sort of OpenStep/NeXTStep based system running here at home... How's that for a short introduction? Grant Bayley __________________________________________________ Grant Bayley AusMac Archive Admin, Mac Evangelist Spare Time Mac Internet Consultant http://mac.unsw.edu.au/ gbayley@geko.net.au __________________________________________________ From joel@yinu.co.jp Tue Mar 18 05:20:43 1997 Return-Path: Received: from ncgw.pro.or.jp by omnigroup.com (NX5.67f2/NX3.0M) id AA12967; Tue, 18 Mar 97 05:20:43 -0800 Received: from yinugw.yinu.co.jp (yinugw.yinu.co.jp [202.234.0.49]) by ncgw.pro.or.jp (8.7.5/3.5Wbeta-97010421) with SMTP id WAA17302 for ; Tue, 18 Mar 1997 22:20:38 +0900 (JST) Received: from woodstock.yinu.co.jp ([192.47.108.3]) by yinugw.yinu.co.jp (8.6.4/3.4Wbeta5-yinugw) with ESMTP id WAA00762 for ; Tue, 18 Mar 1997 22:20:36 +0900 Received: from [192.47.108.36] by woodstock.yinu.co.jp (8.6.4/3.4Wbeta5-yinu_machines) id WAA17485; Tue, 18 Mar 1997 22:20:31 +0900 Message-Id: Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Date: Tue, 18 Mar 1997 22:31:34 +0900 To: rhapsody-dev@omnigroup.com From: joel@yinu.co.jp (Joel Floyd) Subject: Real-Time Services >Mach is not a real-time OS, so the MIDI >timing is _never_ realiable. If you quit all other apps and jack up the MIDI >app's priority, you typically get fine performances even on a 68040, but >even then hiccups are still possible. (It's that "preemptive" part of the OS >getting in the way. This is one of those rare cases where that can cause >problems.) I've had little or no troubles on my Pentium Pro 200 system, >though. And even the 68040 was bearable, so I would expect a PowerPC system >should give acceptable performance. How do you change your applications priority, what about servers do they automatically get a higher priority. I know Windows NT has a bug(?) thing that lets the Application set its priority so that it gets real-time access. I think this would be neat for games. Since we have Unix boxes in the office I know doing real-time stuff is extremely difficult; Another reason I think this would be good, is that when you record a movie, you need all the CPU power you can get to record and compress in real-time. I would also like to know the different states an application can be put in, for instance what about sleep (like on a unix box). etc... Joel From adul@cmg.FCNBD.COM Tue Mar 18 06:40:17 1997 Return-Path: Received: from po-external.FCNBD.COM by omnigroup.com (NX5.67f2/NX3.0M) id AA14110; Tue, 18 Mar 97 06:40:17 -0800 Received: from po-internal.FCNBD.COM (internalhost.FCNBD.COM [147.113.104.10]) by po-external.FCNBD.COM (8.8.5/fcnbd/domain/1.5.1) with ESMTP id IAA28715 for ; Tue, 18 Mar 1997 08:45:37 -0600 (CST) Received: from abacab.cmg.FCNBD.COM (abacab.cmg.FCNBD.COM [147.113.118.227]) by po-internal.FCNBD.COM (8.8.5/fcnbd/internal-domain/1.4.1) with ESMTP id IAA05668 for ; Tue, 18 Mar 1997 08:41:40 -0600 (CST) Received: from falstaff.cmg.FCNBD.COM (adul@falstaff.cmg.FCNBD.COM [147.113.112.147]) by abacab.cmg.FCNBD.COM (8.8.5/fcnbd/server-subdomain/2.3) with ESMTP id IAA26167 for ; Tue, 18 Mar 1997 08:40:10 -0600 (CST) Received: (from adul@localhost) by falstaff.cmg.FCNBD.COM (8.7.3/8.7.1) id IAA08413 for rhapsody-dev@omnigroup.com; Tue, 18 Mar 1997 08:40:09 -0600 (CST) Message-Id: <199703181440.IAA08413@falstaff.cmg.FCNBD.COM> Content-Type: text/plain Mime-Version: 1.0 (NeXT Mail 3.3 v118.2) In-Reply-To: <199703180132.RAA10645@litehouse.eng.sun.com> X-Nextstep-Mailer: Mail 3.3 (Enhance 2.0b5) Received: by NeXT.Mailer (1.118.2) From: Albert Dul Date: Tue, 18 Mar 97 08:40:07 -0600 To: rhapsody-dev@omnigroup.com Subject: Quicky Intro References: <199703180132.RAA10645@litehouse.eng.sun.com> I've been a NextStep developer for ~5 yrs. and have started Mac OS development for my masters thesis. Needless to say, I'm a happy camper, and will crank up development in '98 when I buy a PowerMac to actually run Rhapsody :-). My Q605 won't cut it. Oooooo, Objective-C on a Mac... Albert Dul aldul@concentric.net From bentley@crenelle.com Tue Mar 18 07:05:13 1997 Return-Path: Received: from dns.bl.com by omnigroup.com (NX5.67f2/NX3.0M) id AA14661; Tue, 18 Mar 97 07:05:13 -0800 Received: from [206.185.114.20] (alpha.crenelle.com [206.185.114.20]) by dns.bl.com (8.8.4/1.3) with ESMTP id JAA06713 for ; Tue, 18 Mar 1997 09:05:10 -0600 (CST) X-Sender: bentley@mail.bl.com Message-Id: In-Reply-To: <9703180747.AA09254@omnigroup.com> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Date: Tue, 18 Mar 1997 09:05:13 -0600 To: rhapsody-dev@omnigroup.com From: Michael Brian Bentley Subject: Re: PPC 601 and NuBus based systems Wil sez: >On the other hand, it's worth mentioning that OpenStep/Mach 4.1 already DOES >support the Nubus, so I can't see any technical reason why Rhapsody >couldn't. > >However, there is no doubt it would take some amount of work to support the >Macs particular implementation (and to support it using DriverKit, since the >old Nubus drivers aren't part of DriverKit as far as I know), and I'm sure >right now the porting team is trying to commit to as little work as >possible, >so they can ship on time. I bring up the following because it may relate to the availability of Next support of the NuBUS. There are a number of companies on Mac that rely on the MacOS getting out of the way so that the app gets seamless access to the facilities. Audio and video editing are a couple that have no tolerance for stream discontinuities. Someone mentioned in a job interview the other day that the obvious solution is to pack more buffering hardware on add-on cards and bundle them with shrinkwrap applications. Not having to do that, I guess, has been a MacOS edge for a while. I am hoping that a "quick" release of the Rhapsody prototypes includes a solution for making sure isochronous jobs get enough room without (or even despite!) playing priority games; I envision app developers ceiling the default priority assignments under Mach as a matter of course. It is effectively what is done now in event loop handling and cooperative schedule handling: if you're too egalitarian, the other app will suck the bone marrow out of your share of time. What do audio and video editing tools under OpenStep for Mach do now to prevent stream discontinuities? -m Michael Brian Bentley / bentley@crenelle.com / Crenelle Inc. 1935 West Pratt Blvd Suite 3, Chicago Illinois 60626-3133 Voice (773)-508-9009 Fax (773)-465-2399 Web www.crenelle.com Network Applications Development Specializing in Mac OS From kam@kamit.com Tue Mar 18 07:06:33 1997 Return-Path: Received: from babel.enteract.com by omnigroup.com (NX5.67f2/NX3.0M) id AA14773; Tue, 18 Mar 97 07:06:33 -0800 Received: (qmail 26153 invoked from network); 18 Mar 1997 15:06:29 -0000 Received: from enteract.com (kam@206.54.252.1) by babel.enteract.com with SMTP; 18 Mar 1997 15:06:29 -0000 Date: Tue, 18 Mar 1997 09:06:29 -0600 (CST) From: "Kevin A. Mitchell" X-Sender: kam@enteract.com To: rhapsody-dev@omnigroup.com Subject: Introduction Message-Id: Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII I've been developing and selling the GIFConverter shareware program since 1988. I've been tinkering with Macintosh programming since 1986 or so. I also program VAX/VMS, tons of Unices, and am learning Windows NT (all for my 'day' job). It excites me to find new things to learn, and so I'm eagerly looking forward to working with Rhapsody. I'd like to continue the internal cleanup work I'm doing on GIFConverter, separate the works from the interface, and port it all to OpenStep. As of now, I know nothing about it. And I need to get my hands on some equipment that will run Rhapsody. So there's a quick introduction. I'm in. Friday sucked, and somebody I look up to said "The best thing about the past is it's over." Forward! Kevin A. Mitchell, kam@kamit.com Personal: http://www.kamit.com/ GIFConverter: gifconverter@kamit.com or kam@kagi.com http://www.kamit.com/gifconverter.html From jperry@wizvax.net Tue Mar 18 07:14:44 1997 Return-Path: Received: from wizvax.wizvax.net by omnigroup.com (NX5.67f2/NX3.0M) id AA15104; Tue, 18 Mar 97 07:14:44 -0800 Date: Tue, 18 Mar 97 07:14:44 -0800 Received: (qmail 13548 invoked from network); 18 Mar 1997 15:14:40 -0000 Received: from pt19.wizvax.net (HELO ?204.97.177.19?) (204.97.177.19) by wizvax.wizvax.net with SMTP; 18 Mar 1997 15:14:40 -0000 X-Sender: jperry@mail1.wizvax.net Message-Id: Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" To: rhapsody-dev@omnigroup.com From: jperry@wizvax.net (j perry) Subject: RE: Moving to Rhapsody from Smalltalk I'm another coming from SmalltalkAgents on Mac. STA has something really great called a "block" and it looks something like this: outsidevar := 3; // the block can see this if you want it to myblock := [ insidevar := 5; // look, local vars inside the block beep; // do something ] ensure: [ closemyfile; ] // on exception the block executes this mycollection iterateWithBlock: myblock; Having the outside var used by the block can be a bit dicey (in this case you are giving the block a "context") but it's there if you want it. There are other options besides ensure: to tack onto the block. I've never done this but I think you can make the block a persistant object and store it somewhere if that's appropriate; at any rate the block is itself a first-class object. Is there anything comparable to this in the ObjectiveC language (or extensions in the OpenStep framework)? (Maybe this is a "lambda" variable, I don't really know). TIA, Jeff !!!!!!!!!!!!!!!!!!! , , jperry@wizvax.net . =================== - From marc@htl.com Tue Mar 18 07:20:14 1997 Return-Path: Received: from htl1.htl.com by omnigroup.com (NX5.67f2/NX3.0M) id AA15274; Tue, 18 Mar 97 07:20:14 -0800 Received: from HTL.HTL.COM (MARC.htl.com [199.93.164.110]) by htl1.htl.com (Netscape Mail Server v1.1) with SMTP id AAA72 for ; Tue, 18 Mar 1997 10:31:24 -0500 Message-Id: <2.2.32.19970318151827.00720ae8@htl1.htl.com> X-Sender: marc@htl1.htl.com X-Mailer: Windows Eudora Pro Version 2.2 (32) Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Date: Tue, 18 Mar 1997 10:18:27 -0500 To: Multiple recipients of list From: marc@htl.com (Marc Respass) Subject: Introduction Hi Everyone, Thanks to Wil for hosting this list and for all the valuable info he made available on Semper.Fi. I'm a Mac developer although very much a beginner (one freeware app and one in development so far). I've never used a NeXT system but I did download the docs and I'm eager to get started learning. I work in an all Windows shop (NT servers, 95 clients) doing mostly Paradox 7 database development and some misc. C++ dev. I'm also the webmaster here. I have this crazy idea that Apple will send me Openstep/Mach, I'll be able to do database dev in Openstep and move our web server to Nextstep :). BTW, I live in central Massachusetts. What's the NeXT market like around here? --Marc Respass HTL Technologies, Inc. http://www.htl.com From markm@mail.tyrell.com Tue Mar 18 07:20:19 1997 Return-Path: Received: from apex.tyrell.com by omnigroup.com (NX5.67f2/NX3.0M) id AA15280; Tue, 18 Mar 97 07:20:19 -0800 Received: from [206.170.253.84] (cdungeon.tyrell.com [206.170.253.84]) by mail.tyrell.com (8.6.12/8.6.12-MT2.2) with ESMTP id HAA26295 for ; Tue, 18 Mar 1997 07:19:12 PST (-0800) Message-Id: In-Reply-To: <9703181124.AA29320@misckit.com> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Date: Tue, 18 Mar 1997 07:20:40 -0700 To: rhapsody-dev@omnigroup.com From: "Mark F. Murphy" Subject: Re: MIDI (was Re: Introductions) At 3:29 AM -0800 3/18/97, Donald A. Yacktman wrote: >The big problem is that NeXT's Mach is not a real-time OS, so the MIDI >timing is _never_ realiable. I think Apple will have to address this somehow. QuickTime is based on correct timing. So maybe when they port QTML, there might be some support!? mark --------------------------------------------------------------------------- Mark F. Murphy, Director Software Development (markm@mail.tyrell.com) Tyrell Software Corp (http://www.tyrell.com) --------------------------------------------------------------------------- Member of Families Against Internet Censorship: rainbow.rmi.net/~fagin/faic From andrew Tue Mar 18 07:24:50 1997 Return-Path: Received: from bubastis by omnigroup.com (NX5.67f2/NX3.0M) id AA15483; Tue, 18 Mar 97 07:24:50 -0800 Message-Id: <9703181524.AA15483@omnigroup.com> Received: by bubastis.omnigroup.com (NX5.67g/NX3.0X) id AA11067; Tue, 18 Mar 97 07:25:01 -0800 Content-Type: text/plain Mime-Version: 1.0 (NeXT Mail 4.1mach v148) X-Image-Url: ftp://ftp.omnigroup.com/pub/Images/People/andrew.tiff X-Nextstep-Mailer: Mail 4.1mach (Enhance 2.0b5) Received: by NeXT.Mailer (1.148) From: Andrew Abernathy Date: Tue, 18 Mar 97 07:25:00 -0800 To: rhapsody-dev Subject: Re: reply-to field not set? X-Url: http://www.omnigroup.com/People/andrew/ > > I just noticed that the reply-to field in the header is not getting > > set for any of the messages coming originating from the listserver, > > so my reply's are going to individual people instead of to the > > list... > > That's an intentional setting, the default behavior we've always had on > the other NeXT mailing lists. However, I don't have any strong feelings > one way or the other about the setting, so if people prefer that their > replies go to the list by default (rather than to the author of a > message), let me know. No no no no no no no please lord no... I _hate_ it when I try to reply to some message and end up sending the reply to the entire list. Most of my replies are to individuals, not to the list, and that's something I'd like to encourage in others, so we don't get five different people saying, to the entire list, "where did you hear that?". (Now, this message is special, of course...) Besides, sometimes I'm telling someone something that isn't for general consumption anyway. When I reply to something in my personal mailbox, the reply should go to the originator of the message, not the list. If I want to send to the list, it's easy enough for me to copy the list address out of the message and paste it into the To or Cc field. -andrew From jperry@wizvax.net Tue Mar 18 09:19:46 1997 Return-Path: Received: from wizvax.wizvax.net by omnigroup.com (NX5.67f2/NX3.0M) id AA01135; Tue, 18 Mar 97 09:19:46 -0800 Date: Tue, 18 Mar 97 09:19:46 -0800 Received: (qmail 18920 invoked from network); 18 Mar 1997 17:19:45 -0000 Received: from pt14.wizvax.net (HELO ?204.97.177.19?) (204.97.177.14) by wizvax.wizvax.net with SMTP; 18 Mar 1997 17:19:45 -0000 X-Sender: jperry@mail1.wizvax.net Message-Id: Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" To: rhapsody-dev@omnigroup.com From: jperry@wizvax.net (j perry) Subject: NextStep OS versions? I'm asking around about used Next systems; in one case for example I can get a used NextStation which "include NS 3.3 user/3.2 developer". How far behind the curve are those versions? What would I be missing? Related to that, do I understand right that very few Next shrinkwrap app developers have actually taken their apps to OpenStep (I read that in a recent semper-fi post)? If that's true, is it because (a) the NextStep code runs fine in all cases under OpenStep, (b) adoption of OpenStep is nil at this time, (c) it's way too much trouble to port when so few users have moved to OpenStep, or (d) some combination? Thanks for any rays of light, Jeff !!!!!!!!!!!!!!!!!!! , , jperry@wizvax.net . =================== - From andrew Tue Mar 18 09:34:04 1997 Return-Path: Received: from bubastis by omnigroup.com (NX5.67f2/NX3.0M) id AA01508; Tue, 18 Mar 97 09:34:04 -0800 Message-Id: <9703181734.AA01508@omnigroup.com> Received: by bubastis.omnigroup.com (NX5.67g/NX3.0X) id AA11298; Tue, 18 Mar 97 09:34:18 -0800 Content-Type: text/plain Mime-Version: 1.0 (NeXT Mail 4.1mach v148) X-Image-Url: ftp://ftp.omnigroup.com/pub/Images/People/andrew.tiff X-Nextstep-Mailer: Mail 4.1mach (Enhance 2.0b5) Received: by NeXT.Mailer (1.148) From: Andrew Abernathy Date: Tue, 18 Mar 97 09:34:17 -0800 To: mac-games-dev@solutions.apple.com, semper.fi@solutions.apple.com, rhapsody-dev Subject: Carmack on NeXTstep & Rhapsody X-Url: http://www.omnigroup.com/People/andrew/ John Carmack (the guy who wrote Quake & Doom etc on NeXTstep) just updated his .plan to give his opinion on different platforms. He comments on everything from Plan 9 to BeOS. For those who are interested in the subject of games on Rhapsody, I thought the portion about NeXTstep & Rhapsody might be of interest. (Those interested in the whole thing can finger johnc@idsoftware.com, or can access the finger info from some place like Blue's News ) -andrew > NEXTSTEP > My faviorite environment. NT and linux both have advantages in > some areas, but if they were on equal footing I would choose NEXTSTEP > hands down. It has all the power of unix (there are lots of things I > miss in NT), the best UI (IMHO, of cource), and it just makes sense on > so many more levels than windows. Yes, you can make windows do > anything you want to if you have enough time to beat on it, but you can > come out of it feeling like you just walked through a sewer. > > In the real world, things aren't on equal footing, and I do most of my > work on NT now. I hold out hope that it may not stay that way. If > apple Does The Right Thing with rhapsody, I will be behind them as much > as I can. NEXTSTEP needs a couple things to support games properly > (video mode changing and low level sound access). If apple/next will > provide them, I will personally port our current win32 products over. > > If I can convince apple to do a good hardware accelerated OpenGL in > rhapsody, I would be very likely to give my win NT machine the cold > shoulder and do future development on rhapsody. (I really don't need > Quickdraw3D evangelists preaching to me right now, thank you) From Jeff_Martin@next.com Tue Mar 18 09:38:10 1997 Return-Path: Received: from next.com by omnigroup.com (NX5.67f2/NX3.0M) id AA01656; Tue, 18 Mar 97 09:38:10 -0800 Received: from storm by oz.next.com (NX5.67f1/NeXT0.1-Aleph-bf) id AA27160; Tue, 18 Mar 97 09:38:07 -0800 Message-Id: <9703181738.AA27160@oz.next.com> Received: by storm.next.com (NX5.67g/NX3.0X) id AA04625; Tue, 18 Mar 97 09:38:05 -0800 Mime-Version: 1.0 (NeXT Mail 4.2mach v148) Content-Type: text/enriched; charset=us-ascii Content-Transfer-Encoding: 7bit Content-Disposition: inline Received: by NeXT.Mailer (1.148) From: Jeff Martin Date: Tue, 18 Mar 97 09:38:01 -0800 To: wjs@omnigroup.com Subject: Re: Drawing Directly To Screen Cc: Multiple recipients of list References: <9703180516.AA07226@omnigroup.com> > The biggest limitations of Interceptor right now are: >1) Not supported by NeXT. On the other hand, it's the basis for NEXTIME, so > it ain't going away (although the API may change 100%). > 2) Can't mix PostScript and Interceptor in the same on-screen rectangle 100% > of the time, as far as I know. > 3) You have to write low-level routines to write to every different kind of > framebuffer in the world. Sigh. > 4) You can't dynamically change the screen resolution, so you're stuck with > whatever depth and size the user selected. Ouch. I don't know how much I can say about this, except that each of these points are being addressed. Complete details will be available in May at the World Wide Developer's Conference. Addressing points 1 & 4: Direct framebuffer access is definitely not going away - it is very important to Apple's substantial education, entertainment and multimedia markets. Additionally applications which decide to take over the whole screen will have access to all framebuffer modes available. Addressing points 2 & 3: We have a plan for integrating PostScript and direct device access (this has to be solved for our QuickDraw3D integration). Once this is implemented the problem of the many different frame buffer formats becomes more manageable - simply write a direct path for your code and a path using the DPS image operator (or more correctly the NSBitmapImageRep's -draw method). If you support the framebuffer layout go the direct route. If you don't, go the DPS route. This also prevents your software from breaking when new framebuffer formats become available. Most sprite based animation and all traditional graphics should try to go the DPS route (using PostScript operators and compositing). This buys you the following: WSIWYG, device independence, automatic printing support, the ability to cast the output of your application to other machines and inherent multi-threading between your application logic and graphics output (we demonstrated this internally several years ago by attaining sustained 95% processor utilization of two PPC 601 processors on a prototype machine running BoinkOut in full speed demo mode). For video and 3D, framebuffer access makes a lot of sense - although this will be incorporated automatically when using QuickTime, QuickDraw 3D or the low-level textured triangle routines available in RAVE (which will also buy you automatic hardware acceleration, if available). Hope this helps! ... jeff martin Graphics Group, NeXT, er, that is Apple Software, I mean Computers, Inc. From bentley@crenelle.com Tue Mar 18 10:04:12 1997 Return-Path: Received: from dns.bl.com by omnigroup.com (NX5.67f2/NX3.0M) id AA02130; Tue, 18 Mar 97 10:04:12 -0800 Received: from [206.185.114.20] (alpha.crenelle.com [206.185.114.20]) by dns.bl.com (8.8.4/1.3) with ESMTP id MAA07263 for ; Tue, 18 Mar 1997 12:04:19 -0600 (CST) X-Sender: bentley@mail.bl.com Message-Id: In-Reply-To: <9703181738.AA27160@oz.next.com> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Date: Tue, 18 Mar 1997 12:04:21 -0600 To: rhapsody-dev@omnigroup.com From: Michael Brian Bentley Subject: Re: Drawing Directly To Screen Is there a demo I can run on an Intel Mach Nextstep 3.3 that clearly demonstrates the speed of DPS sprite rendering? I have no reference information to gauge by. I'm trying to translate this sort of thing into frames per second in a 640 x 480 (or larger) area. -m Michael Brian Bentley / bentley@crenelle.com / Crenelle Inc. 1935 West Pratt Blvd Suite 3, Chicago Illinois 60626-3133 Voice (773)-508-9009 Fax (773)-465-2399 Web www.crenelle.com Network Applications Development Specializing in Mac OS From tims@PARADIGM.bc.ca Tue Mar 18 10:15:13 1997 Return-Path: Received: from [207.107.56.5] by omnigroup.com (NX5.67f2/NX3.0M) id AA02331; Tue, 18 Mar 97 10:15:13 -0800 Received: by 2nt.paradigm.bc.ca with SMTP (Microsoft Exchange Server Internet Mail Connector Version 4.0.994.63) id <01BC3385.C67A44E0@2nt.paradigm.bc.ca>; Tue, 18 Mar 1997 10:18:51 -0800 Message-Id: From: Tim Snider To: 'Multiple recipients of list' , "'joel@yinu.co.jp'" Subject: RE: Real-Time Services Date: Tue, 18 Mar 1997 10:18:50 -0800 X-Mailer: Microsoft Exchange Server Internet Mail Connector Version 4.0.994.63 Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Hi everyone, I've just joined in from SemperFi, I'll do an intro in a little while. Nice to see people talking calmly without wailing. ;*} While I'll agree that Mach (version 2.5) is not a realtime system, from what I've read, the OSF MK(version 3.0), which is the foundation of MkLinux, provides realtime services. I understand that Rhapsody will first be implemented on top of the existing 2.5 kernel with some fixes that have been waiting in the wings, after they've gotten everything running I assume that they'll move up to the 3.0 kernel in order to get better realtime services. Has anyone heard yet how they're going to do QuickTime without realtime kernel services? TJ Snider >---------- >From: joel@yinu.co.jp[SMTP:joel@yinu.co.jp] >Sent: Tuesday, March 18, 1997 5:23 AM >To: Multiple recipients of list >Subject: Real-Time Services > > >>Mach is not a real-time OS, so the MIDI >>timing is _never_ realiable. If you quit all other apps and jack up the >>MIDI >>app's priority, you typically get fine performances even on a 68040, but >>even then hiccups are still possible. (It's that "preemptive" part of the >>OS >>getting in the way. This is one of those rare cases where that can cause >>problems.) I've had little or no troubles on my Pentium Pro 200 system, >>though. And even the 68040 was bearable, so I would expect a PowerPC system >>should give acceptable performance. > >How do you change your applications priority, what about servers do they >automatically get a higher priority. I know Windows NT has a bug(?) thing >that lets the Application set its priority so that it gets real-time >access. I think this would be neat for games. Since we have Unix boxes in >the office I know doing real-time stuff is extremely difficult; Another >reason I think this would be good, is that when you record a movie, you >need all the CPU power you can get to record and compress in real-time. I >would also like to know the different states an application can be put in, >for instance what about sleep (like on a unix box). etc... > > >Joel > > > From jkolyer@neptune.esystem.com Tue Mar 18 10:16:36 1997 Return-Path: Received: from mail.esystem.com by omnigroup.com (NX5.67f2/NX3.0M) id AA02364; Tue, 18 Mar 97 10:16:36 -0800 Received: from mtcoak ([192.10.8.40]) by mail.esystem.com (8.7.3/8.7.3) with SMTP id KAA14717 for ; Tue, 18 Mar 1997 10:20:44 -0800 (PST) Received: from lisbon by mtcoak (NX5.67e/1.63) id AA01611; Tue, 18 Mar 97 10:19:34 -0800 Message-Id: <9703181819.AA01611@mtcoak> Received: by lisbon.neptune.esystems.com (NX5.67g/NX3.0X) id AA01413; Tue, 18 Mar 97 10:19:34 -0800 Content-Type: text/plain Mime-Version: 1.0 (NeXT Mail 4.1mach v148) In-Reply-To: <9703180734.AA09020@omnigroup.com> X-Nextstep-Mailer: Mail 3.3 (Enhance 1.3) Received: by NeXT.Mailer (1.148) From: Jonathan Kolyer Date: Tue, 18 Mar 97 10:19:32 -0800 To: rhapsody-dev@omnigroup.com Subject: Re: NextStep vs OpenStep Reply-To: jkolyer@neptune.esystem.com References: <9703180734.AA09020@omnigroup.com> X-Phone: (415) 780 3784 OPENSTEP refers to NeXT's implementation of OpenStep. OPENSTEP for Mach, and OPENSTEP for NT are NeXT products conforming to the OpenStep standard. I don't speak for NeXT (aka Apple), but this is my understanding. JK --------------- Jonathan Kolyer Apple Computer (The company formerly known as "NeXT Software") jkolyer@next.com jkolyer@neptune.esystem.com (on-site) From Robert.Kieffer@Eng.Sun.COM Tue Mar 18 10:25:37 1997 Return-Path: Received: from [192.9.25.5] by omnigroup.com (NX5.67f2/NX3.0M) id AA02722; Tue, 18 Mar 97 10:25:37 -0800 Received: from Eng.Sun.COM ([129.146.1.25]) by venus.Sun.COM (SMI-8.6/mail.byaddr) with SMTP id KAA06834 for ; Tue, 18 Mar 1997 10:25:34 -0800 Received: from litehouse.eng.sun.com by Eng.Sun.COM (SMI-8.6/SMI-5.3) id KAA19088; Tue, 18 Mar 1997 10:25:32 -0800 Received: from lhbodie.Eng.Sun.Com by litehouse.eng.sun.com (SMI-8.6/SMI-SVR4) id KAA24234; Tue, 18 Mar 1997 10:25:32 -0800 Message-Id: <199703181825.KAA24234@litehouse.eng.sun.com> Received: by lhbodie.Eng.Sun.Com (NX5.67g/NX3.0X) id AA01089; Tue, 18 Mar 97 10:26:18 -0800 Content-Type: text/plain Mime-Version: 1.0 (NeXT Mail 4.1mach v148) In-Reply-To: <199703180748.BAA29607@athenet.net> X-Nextstep-Mailer: Mail 4.0 (Enhance 2.0b4) Received: by NeXT.Mailer (1.148) From: Robert Kieffer Date: Tue, 18 Mar 97 10:26:15 -0800 To: rhapsody-dev@omnigroup.com Subject: OpenStep sound support Reply-To: Robert.Kieffer@Eng.Sun.COM References: <199703180748.BAA29607@athenet.net> > I have one question about OpenStep: How is music information > handled (MIDI)? Are there music apps available? I would > imagine Opcode would port OMS (unfortunately). Music? Sound? Not. Actually, this isn't quite true. I believe all of the existing OpenStep implementations (Mach and Intel, by NeXT, and Solaris, by Sun) have sound support. Unfortunately, sound/music has always been one of the weaker areas of OpenStep (e.g. I only get sound on my intel box during a high tide, or three days prior to an earthquake of magnitude 5.5 or higher. ;-) ) There was a music kit available in earlier versions of NeXTSTEP (pre 3.x) which did do MIDI stuff. This has been released into the public domain. I don't believe it's OpenStep compliant though. I don't recall who currently maintains this (Don, didn't you have something to do with this?) NeXT had pretty killer sound support in their early hardare since a nifty DSP was part of the architecture, but as they migrated to Intel et al, this really slipped and is now something that few, if any 3rd party developers make use of. :-( --- Robert Kieffer Phone: (415)524-3238 Lighthouse Design, Ltd. Fax: (415)570-7787 2929 Campus Drive, Suite 101 Email: kieffer@litehouse.eng.sun.com San Mateo, CA 94403 WWW: http://www.lighthouse.com/~kieffer From Jeff_Martin@next.com Tue Mar 18 10:37:15 1997 Return-Path: Received: from next.com by omnigroup.com (NX5.67f2/NX3.0M) id AA02950; Tue, 18 Mar 97 10:37:15 -0800 Received: from storm by oz.next.com (NX5.67f1/NeXT0.1-Aleph-bf) id AA28566; Tue, 18 Mar 97 10:36:34 -0800 Message-Id: <9703181836.AA28566@oz.next.com> Received: by storm.next.com (NX5.67g/NX3.0X) id AA04670; Tue, 18 Mar 97 10:36:33 -0800 Mime-Version: 1.0 (NeXT Mail 4.2mach v148) Content-Type: text/enriched; charset=us-ascii Content-Transfer-Encoding: 7bit Content-Disposition: inline Received: by NeXT.Mailer (1.148) From: Jeff Martin Date: Tue, 18 Mar 97 10:36:30 -0800 To: bentley@crenelle.com Subject: Re: Drawing Directly To Screen Cc: Multiple recipients of list References: > Is there a demo I can run on an Intel Mach Nextstep 3.3 that clearly > demonstrates the speed of DPS sprite rendering? /NextDeveloper/Demos/BoinkOut has a demo mode and can be configured in it's info panel to run full tilt. If you calculate the number of levels completed per minute, these are known as BoinkMarks. :-). A better example is a game called Xox, which is probably available on some ftp site. Both are written by NeXT game guru Sam Streeper. Xox moves around a couple of dozen major sprites plus a star field background on a 1152x832x16bit game field in real time (on my pentium 150). ... jeff From mhoover@minerva.athenet.net Tue Mar 18 10:39:30 1997 Return-Path: Received: from minerva.athenet.net by omnigroup.com (NX5.67f2/NX3.0M) id AA03195; Tue, 18 Mar 97 10:39:30 -0800 Received: from [143.44.248.22] (hooverm.student.lawrence.edu [143.44.248.22]) by athenet.net (8.7.5/8.7.5) with SMTP id MAA06990 for ; Tue, 18 Mar 1997 12:39:25 -0600 (CST) Message-Id: <199703181839.MAA06990@athenet.net> Subject: RE: Real-Time Services Date: Tue, 18 Mar 97 12:45:19 -0600 X-Sender: mhoover@pop.athenet.net X-Mailer: Claris Emailer 1.1 From: Michael Hoover To: "Multiple recipients of list" Mime-Version: 1.0 Content-Type: text/plain; charset="US-ASCII" So does this mean there is no equivalent to the Time Manager? I might be showing my ignorance here, but just because we change from one OS to another does that mean we don't have interrupts anymore? My knowledge of interrupts is very little, just that you can do to-the-millisecond timing things with the Time Manager. Michael Hoover From Jeff_Martin@next.com Tue Mar 18 10:53:34 1997 Return-Path: