shebs@Apple.COM (Stan Shebs) (11/30/90)
After much work and many delays (not all of them technical, ahem), we are very pleased to announce the public release of our port of the Gnu C Compiler (GCC) to MPW*. GCC is a highly optimizing ANSI C compiler; the MPW port is a drop-in replacement for MPW C, handling nearly all of Apple's extensions to C. GCC also includes many new capabilities, such as inline assembly code, aggressive function inlining, and nonconstant initializers. In addition, GCC comes with *complete*, well-structured and well-documented sources! (amazing but true!) MPW GCC is not a replacement for MPW C! You will need to have MPW 3.1 or later, MPW C, and MPW Asm. There is also a script for using GCC with MPW C++. Hardware requirements are a 68020/030+68881 Mac, and at least 2Mb RAM, with 4Mb is preferred. At the moment, MPW GCC is only available via anon ftp from apple.com. Look under pub/gnu.mpw; you have the choice of a complete distribution (about 4.4Mb packed, 8Mb unpacked) or a binary-only version (about 1Mb packed, 2Mb unpacked). Both of these must be transferred in BINARY mode, uncompressed, unbinhexed, and extracted via StuffIt (I used 1.5.1). You can also get MPW GCC from anybody else who has a copy. As per the Gnu Public License, you can copy, modify, and distribute MPW GCC as much as you want; just don't prevent anybody else from doing the same! There are plans afoot to make GCC available on a developer CD, probably within the next few months. This is probably the best approach if you're not in a hurry. If you have any problems or discover any bugs, please let us know about them! The address "mpwgcc@apple.com" will forward to me and other support folks; mailing me directly will also work, but the response will be slower. Incidentally, MPW GCC is *not* a supported product; fortunately, you've got all the sources you need to fix problems... Have fun! Stan Shebs Apple ATG System Software shebs@apple.com *In case anyone didn't know, MPW is a trademark of Apple Computer Inc...
brianj@witsend.cs.umd.edu (Brian Johnson) (12/05/90)
This seems like a terrible shame. I love my mac, and would very much like to have all of my favorite gnu (unix version) utilities. But, I think most of us (and especially apple folk) know the Gnu "boycott apple" position. Are we taking advantage of the Free Software Foundation? I think YES:-( FSF promotes open systems, apple doesn't (I know I've paid my R&D overhead to get the proprietary mac os). It bothers me to see FSF software on apple.com. It's like beating Ghandi with his own stick because you can. Brian Johnson Written in gnumacs, on my mac, wired to a unix box. -- Brian Johnson Computer Science Department brianj@cs.umd.edu University of Maryland (301) 405-2725 College Park, Md 20742
amanda@visix.com (Amanda Walker) (12/06/90)
In article <BRIANJ.90Dec5010246@witsend.cs.umd.edu>, brianj@witsend.cs.umd.edu (Brian Johnson) writes: > It bothers me to see FSF software on apple.com. It's like beating Ghandi with > his own stick because you can. It's more like arguing with Ghandi. Providing free access to GNU software is not exactly beating GNU with a stick, now, is it? Whether or not GNU is pleased is a separate issue. The fact that GNU software can be ported by anyone to any hardware is *not* an accident. It's part and parcel of the GNU approach to software. The GNU boycott (as I have heard it explained) is a request for people not to encourage or support the use GNU software on Apple hardware. This is certainly their prerogative. However, it is also Apple's prerogative to not boycott themselves, and to provide the support and encouragement that the GNU project will not. There's no such thing as a free lunch--if you give something to everyone, that means you may give it to people you disagree with. That's life... They way I see it, everybody wins. The GNU project can continue to refuse to touch the Macintosh, and more people can use GNU software. Whether any given person uses MPW GCC depends on which of these things they think is more important. Seems fair enough to me. If you're going to promote universal access, you must accept the downside of it as well as the benefits. Amanda Walker Visix Software Inc.
russotto@eng.umd.edu (Matthew T. Russotto) (12/06/90)
In article <BRIANJ.90Dec5010246@witsend.cs.umd.edu> brianj@witsend.cs.umd.edu (Brian Johnson) writes: >This seems like a terrible shame. I love my mac, and would very much >like to have all of my favorite gnu (unix version) utilities. But, I >think most of us (and especially apple folk) know the Gnu "boycott >apple" position. > >Are we taking advantage of the Free Software Foundation? I think >YES:-( > >FSF promotes open systems, apple doesn't (I know I've paid my R&D >overhead to get the proprietary mac os). > >It bothers me to see FSF software on apple.com. It's like beating Ghandi with >his own stick because you can. FSF would be trying to have it's cake and eat it too. You can't support the principle that 'software should be free', and then object when someone else takes your software and does something you don't like to it (like making a version for the Evil Company Of The Year). You can't say that "Software should be free, except for supporters of Evil Company", and expect to get any recognition. To it's credit, FSF has not (to my knowledge) tried to stop porting of their software to Apple machines, it has only withheld it's support. (and they don't seem to have 'Support should be free' as part of their philosophy) -- Matthew T. Russotto russotto@eng.umd.edu russotto@wam.umd.edu .sig under construction, like the rest of this campus.
schorsch@oxy.edu (Brent William Schorsch) (12/06/90)
What exactly is GNU??? reply by email if this is known by most people... Thanks! -Brent schorsch (schorsch@oxy.edu)
daven@svc.portal.com (David Newman) (12/07/90)
In article <1990Dec5.202855.6434@eng.umd.edu>, russotto@eng.umd.edu (Matthew T. Russotto) writes: > In article <BRIANJ.90Dec5010246@witsend.cs.umd.edu> brianj@witsend.cs.umd.edu (Brian Johnson) writes: > >Are we taking advantage of the Free Software Foundation? I think > >YES:-( > > > >It bothers me to see FSF software on apple.com. It's like beating Ghandi with > >his own stick because you can. > > FSF would be trying to have it's cake and eat it too. You can't support the > principle that 'software should be free', and then object when someone else > takes your software and does something you don't like to it (like making > a version for the Evil Company Of The Year). You can't say that "Software > should be free, except for supporters of Evil Company", and expect to get > any recognition. Here! Here! I too find FSF's attitude a bit contradictory. -Dave Newman --
laird@chinet.chi.il.us (Laird J. Heal) (12/11/90)
In article <1990Dec5.202855.6434@eng.umd.edu> russotto@eng.umd.edu (Matthew T. Russotto) writes: >In article <BRIANJ.90Dec5010246@witsend.cs.umd.edu> brianj@witsend.cs.umd.edu (Brian Johnson) writes: >> >>It bothers me to see FSF software on apple.com. It's like beating Ghandi with >>his own stick because you can. > >FSF would be trying to have it's cake and eat it too. You can't support the >principle that 'software should be free', and then object when someone else >takes your software and does something you don't like to it (like making >a version for the Evil Company Of The Year). You can't say that "Software >should be free, except for supporters of Evil Company", and expect to get >any recognition. > >To it's credit, FSF has not (to my knowledge) tried to stop porting of their >software to Apple machines, it has only withheld it's support. (and they >don't seem to have 'Support should be free' as part of their philosophy) I've sent the FSF people mail encouraging them to develop a "support is a part of our price" philosophy but their agenda is, purely and simply, that there should be no ownership rights in software. This is a bit extreme, at least from my vantage point of having used software writing to pay quite a few bills in the past few years. My point was that so many companies go out of business (or in Apple's case discontinue support) after a few years and leave their customers high and dry with either the option of reverse-engineering their hardware and software (which Apple's licensing agreements do not exactly encourage but recent court cases seem to be allowing) or of purchasing new systems from whichever vendor is still in business and willing to promise to provide support. Here, however I would like to note that the GNU Public License (which Apple is probably subscribing to - I have not bothered to ftp COPYING from the gcc distribution at apple) requires a vendor to make ALL source code available. Apple is saying that the binaries they are distributing require the sources for gcc to be compiled by the MPW C compiler. It would seem to me, then, that they are incorporating the MPW C compiler as a part of their port of gcc, and they should release the sources to MPW C as an integral part of the distribution. What are others' views on this? I'd like the sources, sure, not to compete but for that time, long from now, when Apple decides that Macintosh and especially MPW no longer fit in with their corporate business plans. FSF would like Apple to release the sources, per their agreement, scot free to whomever wants them. -- Laird J. Heal The Usenet is dead! Home: laird@slum.mv.com (Salem, NH) Long Live the Usenet! Away: laird@chinet.chi.il.us
shebs@Apple.COM (Stan Shebs) (12/12/90)
In article <1990Dec11.140812.1003@chinet.chi.il.us> laird@chinet.chi.il.us (Laird J. Heal) writes: >Apple is saying that the binaries they are distributing require the sources >for gcc to be compiled by the MPW C compiler. GCC can recompile itself, but we include builds with MPW C as a convenience; it's not required. >It would seem to me, then, >that they are incorporating the MPW C compiler as a part of their port of >gcc, and they should release the sources to MPW C as an integral part of >the distribution. Following this line of reasoning, "complete" sources includes libs, includes, the command shell, the operating system, the test code used to exercise the motherboard in the factory, ad infinitum. Pretty silly. Note that GCC-dependent vendors such as NeXT haven't volunteered to release all of their sources either... Stan Shebs Apple ATG System Software shebs@apple.com
anders@verity.com (Anders Wallgren) (12/12/90)
In article <1990Dec11.140812.1003@chinet.chi.il.us>, laird@chinet (Laird J. Heal) writes: > >Here, however I would like to note that the GNU Public License (which Apple >is probably subscribing to - I have not bothered to ftp COPYING from the gcc >distribution at apple) requires a vendor to make ALL source code available. ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ >Apple is saying that the binaries they are distributing require the sources >for gcc to be compiled by the MPW C compiler. It would seem to me, then, >that they are incorporating the MPW C compiler as a part of their port of >gcc, and they should release the sources to MPW C as an integral part of >the distribution. > This is wrong and just plain absurd! Along that same line of reasoning, since we need MPW in the first place, Apple should release the source code to that as well. Oh, but you also need to run the Macintosh OS in order to run MPW, so I guess we'll have to give you the sources to that, too. I don't think this was the original intention of the GNU project's copyleft policy. More specifically, at the bottom of Paragraph 3 in COPYING: "Source code for a work means the preferred form of the work for making modifications to it. For an executable file, complete source code means all the source code for all modules it contains; but, as a special exception, it need not include source code for modules which are standard libraries that accompany the operating system on which the executable file runs, or for standard header files or definitions files that accompany the operating system." The key phrase is "for making modifications to it," not "for compiling it in the first place." I interpret this to also include tools and executables which accompany the operating system, namely MPW and it's C compiler. Don't get me wrong - I think the GNU project has contributed immensely to the state of the art, and I believe that its policy on distributing source code works well and that GNU are well within their rights to do so. However, I don't think its intention was, nor is it capable of, being a means of coercing other entities (ie. Apple, or you and me) to follow the same philosophy as the GNU project with respect to THEIR PRODUCTS. I leave the argument of whether it is morally/ethically correct to port and support GNU products on platforms that GNU is boycotting up to each individual. Personally, I don't think there's anything wrong with it as long as the source code policy is followed, but I can understand why other people argue differently.
CXT105@psuvm.psu.edu (Christopher Tate) (12/12/90)
The next logical question here, I think, is....
Can we expect to see g++ (the GNU version of C++) ported, also?
-------
Christopher Tate | "Living in a fisheye lens, caught
| in the camera eye; I have no
cxt105@psuvm.psu.edu | heart to lie: I can't pretend a
{...}!psuvax1!psuvm.bitnet!cxt105 | stranger is a long-awaited friend."
cxt105@psuvm.bitnet | -- Rush, "Limelight"dorner@pequod.cso.uiuc.edu (Steve Dorner) (12/12/90)
>GNU boycott (as I have heard it explained) is a request for people not to >encourage or support the use GNU software on Apple hardware. It's more than that. Stallman wants no one to buy or use any Apple equipment for any purpose. If Apple has what's best for you, he suggests you wait until somebody else provides it. He also wants everyone to boycott NeXT until the offer for a free copy of Lotus' spreadsheet expires. So far as I know, he has nothing against Decwriters and PDP-11's. Free software is fine (I produce some). But somebody's gotta pay the bills, too, and I think it's very naive to expect lots of good software to be written without compensation. I think that's called "communism", and I don't think anyone anywhere has made it work. I also think it's ridiculous to try to copyright a trash can. -- -- Steve Dorner, U of Illinois Computing Services Office Internet: s-dorner@uiuc.edu UUCP: uunet!uiucuxc!uiuc.edu!s-dorner
lyman@headcrash.Berkeley.EDU (Lyman S. Taylor) (12/12/90)
In article <1990Dec11.140812.1003@chinet.chi.il.us> laird@chinet.chi.il.us (Laird J. Heal) writes: > >Here, however I would like to note that the GNU Public License (which Apple >is probably subscribing to - I have not bothered to ftp COPYING from the gcc >distribution at apple) requires a vendor to make ALL source code available. >Apple is saying that the binaries they are distributing require the sources >for gcc to be compiled by the MPW C compiler. It would seem to me, then, >that they are incorporating the MPW C compiler as a part of their port of >gcc, and they should release the sources to MPW C as an integral part of >the distribution. I think you should carefully read the file COPYING before making statements as to what it does and does not say. This has got to be one of the most misread documents I know of ( mainly because people read what the want to see. It's actually a straightforward document.. as legal documents go :-) ). No, I'm an not a lawyer and I do not think you have to be a lawyer to understand the License. COPYING ( for those who don't know it is the part of the GNU Public License that gives some people "fits" ) would require in this case, since gcc is the program in question, that all the code, including the mods, to gcc be made available. Passing this code through a "tool", in this case MPW C, does not free the MPW C source. People seem to get the wrong impression that the GNU Public License "frees" everything it touches. It does NOT. Using gnu-emacs to edit your source code does not free your code. Since the license covers all FSF "tools" there was ( and I guess, still is ) some confusion about the use of gcc since it includes header files. This is why there's the special exemption clause about header files in COPYING. Therefore, using gcc to compile your code does NOT free your code, anymore than eamcs would. Taking the FSF code and incorporating it into your program or modifying the code and redistributing it are a different story. If you want incorporte then its "lawyer time" since this is where copyrights get fuzzy. If you modify and redistribute then, this is a clear violation. I would also like comment on this thread of: "What does Apple mean by doing this port.... " I would like to than Stan for doing the port. ( Especially since he probably knew that he see responses along these lines ). I really doubt this was an "Apple" project. However it's nice to know that Apple would help "deliver" a nice tool to have once it has been created. Lyman
francis@daisy.uchicago.edu (Francis Stracke) (12/12/90)
In article <1990Dec11.210748.29010@ux1.cso.uiuc.edu> dorner@pequod.cso.uiuc.edu (Steve Dorner) writes: >>GNU boycott (as I have heard it explained) is a request for people not to >>encourage or support the use GNU software on Apple hardware. > >It's more than that. Stallman wants no one to buy or use any Apple equipment >for any purpose. If Apple has what's best for you, he suggests you wait >until somebody else provides it. > >He also wants everyone to boycott NeXT until the offer for a free copy of >Lotus' spreadsheet expires. > No offense to anybody at FSF, but this sounds a little silly. Macs are one of the best things that could possibly have happened for the computer world as a whole. I really appreciate GNU stuff. I'm using GNUmacs under X now, with at least some mouse support, & it's very nice--FOR A UNIX PRODUCT. If anyone suggested to me that the average person could use it quickly and easily, I'd laugh in his face. I've taught people to use a Mac word processor--include fonts, style, and footnotes-- in under half an hour. These are not computer people; these are people who've never used any keyboard with even a minimal brain attached. Moreover, I've just taught them how to use, at least partially, *every* Mac word processor there is. They love it! (I love it! :-) These people go on to buy their own Macs, get software for them, etc. Remember: free software means nothing to someone who doesn't have a computer. >I also think it's ridiculous to try to copyright a trash can. Ditto. Or to sue for copying what you yourself copied, or to sue for such awful copying as Microsoft had done at the time Apple started their lawsuit. (I used Windows about that time, & I *could not* figure out how to resize a window left-right. No icon interface in their Finder-equivalent, either.) Comment from David Letterman on the Xerox suit: "Xerox sues somebody for copying?" | Francis Stracke | My opinions are my own. I don't steal them.| | Department of Mathematics |=============================================| | University of Chicago | Until you stalk and overrun, | | francis@zaphod.uchicago.edu | you can't devour anyone. -- Hobbes |
russotto@eng.umd.edu (Matthew T. Russotto) (12/12/90)
On a related note-- will GCC work with MPW 2.02? I'm happy with MPW 2.02
(and, more relevantly, don't have the $$$ for 3.1-- and am not sure I would
want to anyway, after reading the horror file of bugs in 3.0), but GCC would
be an interesting addition.
--
Matthew T. Russotto russotto@eng.umd.edu russotto@wam.umd.edu
.sig under construction, like the rest of this campus.russotto@eng.umd.edu (Matthew T. Russotto) (12/12/90)
In article <1990Dec11.210748.29010@ux1.cso.uiuc.edu> dorner@pequod.cso.uiuc.edu (Steve Dorner) writes: >>GNU boycott (as I have heard it explained) is a request for people not to >>encourage or support the use GNU software on Apple hardware. > >It's more than that. Stallman wants no one to buy or use any Apple equipment >for any purpose. If Apple has what's best for you, he suggests you wait >until somebody else provides it. > >He also wants everyone to boycott NeXT until the offer for a free copy of >Lotus' spreadsheet expires. Yawn. If Stallman was to boycott every company which had something to do with a company that did things he didn't like, he would quickly find he had no equipment he could use. >Free software is fine (I produce some). But somebody's gotta pay the bills, >too, and I think it's very naive to expect lots of good software to be written >without compensation. I think that's called "communism", and I don't think >anyone anywhere has made it work. No, communism is when you put a gun to someones head and force them to write software without compensation above that needed to keep you alive. I think rms et al have shown that lots of good software can be written without compensation, though I'd love to know how they do it... -- Matthew T. Russotto russotto@eng.umd.edu russotto@wam.umd.edu .sig under construction, like the rest of this campus.
shebs@Apple.COM (Stan Shebs) (12/13/90)
In article <90345.160554CXT105@psuvm.psu.edu> CXT105@psuvm.psu.edu (Christopher Tate) writes: > >Can we expect to see g++ (the GNU version of C++) ported, also? Not by me! 1/2 :-) Seriously, the MPW CFront is more up-to-date than G++. The next major release of GCC, 2.0, is supposed to have C++ support built in. *No* promises about when FSF will release it, but I'll probably try to incorporate MPW changes as soon as I get my hands on a copy. Stan Shebs Apple ATG System Software shebs@apple.com
beard@ux5.lbl.gov (Patrick C Beard) (12/13/90)
In article <tv0jnrrhma@visix.com> amanda@visix.com (Amanda Walker) writes: #In article <BRIANJ.90Dec5010246@witsend.cs.umd.edu>, #brianj@witsend.cs.umd.edu (Brian Johnson) writes: #> It bothers me to see FSF software on apple.com. It's like beating Ghandi with #> his own stick because you can. # #It's more like arguing with Ghandi. Providing free access to GNU software #is not exactly beating GNU with a stick, now, is it? Whether or not GNU #is pleased is a separate issue. Exactly. What would you rather see, Apple keep MPW GCC to themselves? That would be software hoarding. I would say that Apple in some small way is going along with FSF's intentions. I think it definitely benefits all of us. -- ------------------------------------------------------------------------------- - Patrick Beard, Macintosh Programmer (beard@lbl.gov) - - Berkeley Systems, Inc. ".......<dead air>.......Good day!" - Paul Harvey - -------------------------------------------------------------------------------
laird@chinet.chi.il.us (Laird J. Heal) (12/13/90)
In article <11494@goofy.Apple.COM> shebs@Apple.COM (Stan Shebs) writes: >In article <1990Dec11.140812.1003@chinet.chi.il.us> laird@chinet.chi.il.us (Laird J. Heal) writes: > >>Apple is saying that the binaries they are distributing require the sources >>for gcc to be compiled by the MPW C compiler. > >GCC can recompile itself, but we include builds with MPW C as a convenience; >it's not required. I thought I read that there was a nasty bug (SANE on a 68000 - gee that includes my Portable I'm typing on) that necessitated compiling with MPW. Hence, one cannot reproduce his binaries without MPW C. > >>It would seem to me, then, >>that they are incorporating the MPW C compiler as a part of their port of >>gcc, and they should release the sources to MPW C as an integral part of >>the distribution. > >Following this line of reasoning, "complete" sources includes libs, includes, >the command shell, the operating system, the test code used to exercise >the motherboard in the factory, ad infinitum. Pretty silly. Note that >GCC-dependent vendors such as NeXT haven't volunteered to release all of >their sources either... It is a special case of distributing a compiler. Macintosh software are also special cases in that the operating system is subordinate to the particular application. Here, Apple must claim that the gcc complier is a complete package, and the question is where should the line be drawn. It seems to me that a compiler that will not compile itself correctly is not a complete package. It is also true that distributing object-only is mostly a convenience to the vendor. A commercial convenience, but nonetheless a nuisance to the customers. You know how nice it is to look up the PL/S source in the fiche sometimes? I haven't had to do it for years now, but it sure came in handy when I did have to. > > Stan Shebs > Apple ATG System Software > shebs@apple.com I sincerely hope there comes a time when a user of software has the right to the source code, if only when a vendor decides to desupport it. There is no need to compete as a software supplier, but when a commercial user loses support, he should have the right to second-source without the need to have someone reverse-engineer his current system. -- Laird J. Heal The Usenet is dead! Here: laird@chinet.chi.il.us Long Live the Usenet!
laird@chinet.chi.il.us (Laird J. Heal) (12/13/90)
In article <1990Dec11.202347.26210@verity.com> anders@verity.com (Anders Wallgren) writes: >In article <1990Dec11.140812.1003@chinet.chi.il.us>, laird@chinet (Laird J. Heal) writes: >> >>Here, however I would like to note that the GNU Public License (which Apple >>is probably subscribing to - I have not bothered to ftp COPYING from the gcc >>distribution at apple) requires a vendor to make ALL source code available. > ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ >This is wrong and just plain absurd! Along that same line of >reasoning, since we need MPW in the first place, Apple should release >the source code to that as well. You can run an MPW tool standalone. I hope you just got caught up in the spirit of argumentativeness. > Oh, but you also need to run the >Macintosh OS in order to run MPW, so I guess we'll have to give you >the sources to that, too. I don't think this was the original >intention of the GNU project's copyleft policy. More specifically, at >the bottom of Paragraph 3 in COPYING: > > "Source code for a work means the preferred form of the work > for making modifications to it. > >The key phrase is "for making modifications to it," > I believe it was made clear that there was a nasty bug in the compiler which prevented it from being usable if it were compiled by itself. That is, it was only usable if produced by MPW C. That being the case, where does the C compiler begin and where does it end? Standard libraries are one thing, usable source code is another. If the fix to the bug in source code lies in other source code...should the license not cover that other source code? -- Laird J. Heal The Usenet is dead! Here: laird@chinet.chi.il.us Long Live the Usenet!
shebs@Apple.COM (Stan Shebs) (12/14/90)
In article <1990Dec12.154308.5819@eng.umd.edu> russotto@eng.umd.edu (Matthew T. Russotto) writes: >On a related note-- will GCC work with MPW 2.02? <shudder> I don't know, but 3.0 changed many things, probably including many of the assumptions that GCC depends on. Try it and let us know! Stan Shebs Apple ATG System Software shebs@apple.com
shebs@Apple.COM (Stan Shebs) (12/14/90)
In article <1990Dec12.154832.5894@eng.umd.edu> russotto@eng.umd.edu (Matthew T. Russotto) writes: >I think rms et al have shown that lots of good software can be written without >compensation, though I'd love to know how they do it... The general idea is that you get paid to write software that someone needs and is willing to pay for anyway, but doesn't care about trying to make a profit from it. This is commonplace in universities and in DoD+contractor environments, so the bottom line is that your state and federal taxes support GNU. (Note that the GNU Manifesto proposes a national Software Tax, although given the govt's record on using tax money, I suspect that the proceeds would go to building faster Unix kernels rather than some crazy notion like a friendly user interface that non-programmers could use! Gasp! The very idea!) Stan Shebs Apple ATG System Software shebs@apple.com
francis@arthur.uchicago.edu (Francis Stracke) (12/14/90)
In article <1990Dec13.104620.6232@chinet.chi.il.us> laird@chinet.chi.il.us (Laird J. Heal) writes: [stuff about Apple releasing MPW source] >>The key phrase is "for making modifications to it," >> >I believe it was made clear that there was a nasty bug in the compiler >which prevented it from being usable if it were compiled by itself. >That is, it was only usable if produced by MPW C. That being the case, >where does the C compiler begin and where does it end? Standard libraries >are one thing, usable source code is another. If the fix to the bug in >source code lies in other source code...should the license not cover that >other source code? The ROMs are a standard library, you know. We know that many are buggy. Should Apple then release the ROM source? | Francis Stracke | My opinions are my own. I don't steal them.| | Department of Mathematics |=============================================| | University of Chicago | Until you stalk and overrun, | | francis@zaphod.uchicago.edu | you can't devour anyone. -- Hobbes |
jeffh@HyperMail.apple.com (Jeff Holcomb) (12/14/90)
In article <1990Dec13.104024.6150@chinet.chi.il.us> laird@chinet.chi.il.us (Laird J. Heal) writes: > I thought I read that there was a nasty bug (SANE on a 68000 - gee that > includes my Portable I'm typing on) that necessitated compiling with MPW. > Hence, one cannot reproduce his binaries without MPW C. > It is a special case of distributing a compiler. Macintosh software are > also special cases in that the operating system is subordinate to the > particular application. Here, Apple must claim that the gcc complier is > a complete package, and the question is where should the line be drawn. > It seems to me that a compiler that will not compile itself correctly is > not a complete package. The cpp and cc1 files that make up the compiler are the third generation. That is: generation one was created by compiling the sources through MPW C. Generation two was created by compiling the sources using the generation one cpp and cc1, and generation three was created by using the generation two cpp and cc1. MPWGCC is perfectly capable of compiling itself (better than MPW C actually). All Apple has to do to distribute MPWGCC is to provide copies of the source code of MPWGCC. We don't have to offer support, promise updates, etc. As we state in the Read Me file: "This is not an official product of Apple Computer! MPW* GCC was developed to support Advanced Technology Group research efforts, and was deemed sufficiently useful to be made available to a wider audience. If you use this compiler, you should be prepared to support it yourself (full sources are included)." ___________________________________________________________________________ Jeff Holcomb Internet: jeffh@HyperMail.apple.com AppleLink: jeffh@HyperMail.apple.com@INTERNET# GEnie: A2.JEFFH Contracting for Apple ATG Voice: (408) 974-0841 My opinions are not necessarily those of Apple. :-P ___________________________________________________________________________
anders@verity.com (Anders Wallgren) (12/15/90)
In article <1990Dec13.104620.6232@chinet.chi.il.us>, laird@chinet (Laird J. Heal) writes: >In article <1990Dec11.202347.26210@verity.com> anders@verity.com (Anders Wallgren) writes: >>In article <1990Dec11.140812.1003@chinet.chi.il.us>, laird@chinet (Laird J. Heal) writes: >>> >>>Here, however I would like to note that the GNU Public License (which Apple >>>is probably subscribing to - I have not bothered to ftp COPYING from the gcc >>>distribution at apple) requires a vendor to make ALL source code available. >> ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ >>This is wrong and just plain absurd! Along that same line of >>reasoning, since we need MPW in the first place, Apple should release >>the source code to that as well. > >You can run an MPW tool standalone. I hope you just got caught up in >the spirit of argumentativeness. > Huh? I've never been able to run MPW tools without MPW... Maybe I'm missing something, but I don't think this point is at all relevant to the discussion. And no, I'm not being argumentative - the GNU COPYING policy does NOT cover source code of tools used to bring up GNU products on a particular platform. You always need the native tool at least once, even if you are later able to recompile GCC with itself. Think about the GNU products that aren't able to re-generate themselves after they've been compiled the first time. In this case you're not able to complete this last (optional) step, but I don't see what difference that makes. >> Oh, but you also need to run the >>Macintosh OS in order to run MPW, so I guess we'll have to give you >>the sources to that, too. I don't think this was the original >>intention of the GNU project's copyleft policy. More specifically, at >>the bottom of Paragraph 3 in COPYING: >> >> "Source code for a work means the preferred form of the work >> for making modifications to it. >> >>The key phrase is "for making modifications to it," >> >I believe it was made clear that there was a nasty bug in the compiler >which prevented it from being usable if it were compiled by itself. >That is, it was only usable if produced by MPW C. That being the case, >where does the C compiler begin and where does it end? Standard libraries >are one thing, usable source code is another. If the fix to the bug in >source code lies in other source code...should the license not cover that >other source code? No. It doesn't make one whit of difference that GCC is only usable when produced by MPW C. You can't use the fact that there exists a bug in an environment or in modifications to GNU code to make the argument that proprietary source code should be made available. Just so I won't be mistaken for a GNU-basher I'll repeat again what I said at the end of my last message: >Don't get me wrong - I think the GNU project has contributed immensely >to the state of the art, and I believe that its policy on distributing >source code works well and that GNU are well within their rights to do >so. However, I don't think its intention was, nor is it capable of, >being a means of coercing other entities (ie. Apple, or you and me) >to follow the same philosophy as the GNU project with respect to THEIR >PRODUCTS. >-- >Laird J. Heal The Usenet is dead! >Here: laird@chinet.chi.il.us Long Live the Usenet!
Bruce.Hoult@bbs.actrix.gen.nz (12/16/90)
jeffh@HyperMail.apple.com (Jeff Holcomb) writes: >The cpp and cc1 files that make up the compiler are the third generation. >That is: generation one was created by compiling the sources through MPW >C. Generation two was created by compiling the sources using the >generation one cpp and cc1, and generation three was created by using the >generation two cpp and cc1. MPWGCC is perfectly capable of compiling >itself (better than MPW C actually). Call me confused, but shouldn't the 2nd and 3rd generation versions of cpp and cc1 be byte-for-byte identical? (as would all subsequent generations) -- Bruce.Hoult@bbs.actrix.gen.nz Twisted pair: +64 4 772 116 BIX: brucehoult Last Resort: PO Box 4145 Wellington, NZ
Lawson.English@p88.f15.n300.z1.fidonet.org (Lawson English) (12/16/90)
Laird J. Heal writes in a message to All
LJH> Here, however I would like to note that the GNU Public License
LJH> (which Apple is probably subscribing to - I have not bothered
LJH> to ftp COPYING from the gcc distribution at apple) requires a
LJH> vendor to make ALL source code available. Apple is saying that
LJH> the binaries they are distributing require the sources for gcc
LJH> to be compiled by the MPW C compiler. It would seem to me, then,
LJH> that they are incorporating the MPW C compiler as a part of their
LJH> port of gcc, and they should release the sources to MPW C as
LJH> an integral part of the distribution.
As I haven't seen the Apple port, I can only intuit that Apple meant that
their
version only compiles under MPW C. This is no different than FreeSoft's
observation
of which compilers their code compiles under without modification. Sound's
like
you just want to look at Apple's sources... (I'm sure that can be arranged:
Why don't you leave a message here for "GNU Prometheus" and see who
responds...)
Lawson (the above was meant sarcastically, BTW)
--
Uucp: ...{gatech,ames,rutgers}!ncar!asuvax!stjhmc!300!15.88!Lawson.English
Internet: Lawson.English@p88.f15.n300.z1.fidonet.orgjeffh@HyperMail.apple.com (Jeff Holcomb) (12/18/90)
In article <1990Dec16.111012.4314@actrix.gen.nz> Bruce.Hoult@bbs.actrix.gen.nz writes: > Call me confused, but shouldn't the 2nd and 3rd generation versions of cpp > and cc1 be byte-for-byte identical? (as would all subsequent generations) Normally they should match, in fact generation 2 and 3 cpp do. cc1 on the other hand is different. We ran into a few problems getting certain sections of the code through MPW C (it totally lost it). The best way we found was to not compile those sections (mainly loop optimization) when building generation 1. Generation 2 is then built without the loop optimization, though it does compile the source code that was left out previously. Generation 3 is then built with the loop optimization. The difference in code size is about 9k between stage 2 and 3. If you build a generation 4, it does match exactly. ___________________________________________________________________________ Jeff Holcomb Internet: jeffh@HyperMail.apple.com AppleLink: jeffh@HyperMail.apple.com@INTERNET# GEnie: A2.JEFFH Contracting for Apple ATG Voice: (408) 974-0841 My opinions are not necessarily those of Apple. :-P ___________________________________________________________________________ electively, at any time they feel like. They can steal programs off your disk. They can survey your disk to see what programs are on it, and then, for example, sell your "user profile" to advertisers who will flood you with junk mail. Etc. Using Prodigy is exactly like using a Unix system and turning off ALL forms of protection - let anybody logged on do anything they wish. If I connect my PC to some other computer, I want to be very sure that there is some security. Right now I am running Telnet on my PC. There is no way to log in from the outside. You can, if you wish, right now, do ftp to my PC. But you need the password. I have no idea how good the security is but at least it exists. Doug McDonald