rms@AI.MIT.EDU (06/03/89)
1. My lawyer says we can enforce the copyleft. There is no reason to believe a government agency merely by deciding to do so could defeat it. (Aside from that, government agencies tend to like the idea.) 2. The reasons you give for which certain companies won't use GCC seem rather irrational. My policies have been consistent and have varied only in small details ever since the GNU project has existed. 3. I know for a fact that Digital employees use GNU software all over the place; and IBM people have also submitted extensions to it. So I doubt very much that these companies have a uniform policy of the sort you describe. In addition, various companies, including MCC and NeXT, have found in practice that the problems referred to do not exist. 4. It may still be true that some parts of those companies have the beliefs and follow the policies you quote. If so, it's their loss. 5. I am willing to believe that, for this reason, you wish for a compiler that these companies could use in proprietary software. In a nutshell, those parts of those companies demand your support for their policies of restricting the users, and you wish to give it to them. 6. However, I have no interest in helping to develop such a thing. My purpose in developing software is to discourage the practice of forbidding end users to share and modify software. If I had not used the copyleft, I would have entirely failed to advance this aim. 7. I appreciate the help you have given me in testing GCC in the past. However, I don't find it so vital that I would start working toward an end that I don't consider worthwhile, in order to have your cooperation in achieving it. That would be putting the means above the ends. 8. Getting people research funding from DEC or IBM, while I see nothing objectionable about it, is not my aim. I would not be willing to take off the copyleft even to get funds for myself. So it would be silly for me to do so to help you get funds. Accepting the world "as it is" means abandoning the attempt to improve it. If I were willing to accept proprietary software, why not also accept today's state of the art?
spaf@CS.PURDUE.EDU (Gene Spafford) (06/03/89)
Sigh. I'm not getting across to the GNUers with this, I guess. I have heard from people at nearly 16 universities and 4 research labs who do understand, however, and would be interested in a true public domain ANSI C compiler. So, you may not understand (or agree with) what I've been saying, but there does seem to be a group who do...and that's just the response from those on this mailing list! Since this group is for technical discussion about the compiler, and since I don't seem to be getting through, I won't continue posting this here. I would like to respond to a few of your points, though. >> Date: Fri, 2 Jun 89 17:09:00 EDT >> From: rms@ai.mit.edu >> To: spaf >> To: info-gcc@prep.ai.mit.edu >> Subject: Some people won't use GCC >> >> 1. My lawyer says we can enforce the copyleft. There is no reason to >> believe a government agency merely by deciding to do so could defeat >> it. I hope your lawyer is a good one if it ever comes to that. I've talked to a number of lawyers, including those for the University and the general consensus is the copyleft is unenforceable. First and foremost, the whole issue of "shrinkwrap" forms of licenses has not been fully addressed in court. You do have copyright, but it may end at the traditional rights. There's also the question of whether its a valid license since there is no exchange of considerations. Even if the copyleft is ever ruled valid, it is not valid to cause someone to give up their Federal rights under copyright in software they have not yet even written yet (I am told). Of course, unless it comes to trial, we'll never know, but there is enough difference of opinion to give one less than a certain feeling. And if the oppposing party happened to be a government agency....how long can you continue to pay your lawyer? >> 2. The reasons you give for which certain companies won't use GCC seem >> rather irrational. My policies have been consistent and have varied >> only in small details ever since the GNU project has existed. Maybe so, but you are espousing something contrary to their commonly accepted standards of law and business, and you have a reputation as a fanatic. You are not trusted to behave rationally, and so they have to react according to worst-case scenerios. Thus, they aren't being irrational at all. >> 3. I know for a fact that Digital employees use GNU software all over >> the place; and IBM people have also submitted extensions to it. So I >> doubt very much that these companies have a uniform policy of the sort >> you describe. I agree. I also know many of these same people. However, the folks in the areas developing language-based software tools are the ones I've been talking to. And just because employees use it doesn't meant the company as a whole has a policy on it. Yet. If one of those employees writes computer viruses at work, does that mean the company has an official policy allowing it? >> In addition, various companies, including MCC and NeXT, have found in >> practice that the problems referred to do not exist. So? We're not a consortium like MCC nor a start-up like NeXT. Their concerns and risks are different. >> 4. It may still be true that some parts of those companies have the >> beliefs and follow the policies you quote. If so, it's their loss. The whole community's loss, actually, and you could affect it if you wanted (for the better). >> 5. I am willing to believe that, for this reason, you wish for a >> compiler that these companies could use in proprietary software. In a >> nutshell, those parts of those companies demand your support for their >> policies of restricting the users, and you wish to give it to them. Bullshit. You completely misunderstand the interest. I want a compiler that I can use to prototype and demonstrate new software tools and approaches that can then be shared by the community. What I produce will not be marketable. We don't do company proprietary research here -- this is a university! I'm not seeking to oppress the masses or any other such nonsense. I'm attempting to advance the state-of-the-art in software engineering. When I do, I will publish the results and share the technology when I can -- and I won't restrict others in the way it can be used. My bottom line is that GNU software *is* *proprietary* too!! It cannot be used freely by others. It is effectively share-ware -- to use it, we have to support your political philosophy. Calling the software "free" under these circumstances is an interesting use of doublespeak. I don't wish you ill, and I will continue to be interested in what the FSF does. However, I also wish you would really live up to the name "Free Software Foundation."
schmidt@ics.uci.edu (Doug Schmidt) (06/03/89)
In article <8906022354.AA28510@uther.cs.purdue.edu> spaf@CS.PURDUE.EDU (Gene Spafford) writes: ++ Sigh. I'm not getting across to the GNUers with this, I guess. No Gene, your message is coming across loud and clear, like a siren's song. You want us (i.e., supporters of Free Software) to sell-out our morals, to prostitute our beliefs in order to satisfy the status quo, to prostrate ourselves before the altar of mediocrity and greed. You want us to recant our faith that GNU supports a purpose more noble than simply pandering to your `sponsors' demands. You want us to cringe in fear upon hearing that some powerful organizations don't approve of our convictions. You want us to keep our place, and not protest against myopic policies of computer vendors that try robbing our dignity. We understand the song you sing, because we've heard it sung many times before. Whether it be Tiananmen Square or Panama City or Pretoria we understand that those with a vested interest in the status quo will stop at nothing to retain their power and control. If refusing to heed your siren's song makes me a fanatic, then I'm proud to wear that label. Douglas C. Schmidt -- schmidt@ics.uci.edu (ARPA) | Per me si va nella citta' dolente. office: (714) 856-4043 | Per me si va nell'eterno dolore. | Per me si va tra la perduta gente. | Lasciate ogni speranza o voi ch'entrate.
markv@tillamook.uucp (Mark VandeWettering) (06/05/89)
In article <16505@paris.ics.uci.edu> Doug Schmidt <schmidt@zola.ics.uci.edu> writes: >In article <8906022354.AA28510@uther.cs.purdue.edu> spaf@CS.PURDUE.EDU (Gene Spafford) writes: >++ Sigh. I'm not getting across to the GNUers with this, I guess. >No Gene, your message is coming across loud and clear, like a siren's >song. [ Poetic drivel deleted to save on net.alka.seltzer] The Free Software Foundation was founded to make software available to people everywhere, to create software that everyone could use. Unfortunately, the FSF has decided to copyright its works under a very RESTRICTIVE license. All the people who are part of the Free Software Foundation will disagree, claiming that this license is what guarantees that the software will remain free. But you don't need a license to do that. It's called PUBLIC DOMAIN. But, you claim, then people could snarf it up, make proprietary changes, sell it, restrict access to binaries. Yes. But you would always be able to compete with that because your intent was to make it freely available. I admire the Free Software Foundation. Their nearly endless energy and talent has produced some very fine software, and I probably will continue to use their products. However, I have begun to view the "copyleft" under which gnu products are released as being too restrictive. >You want us to recant our faith that GNU supports a purpose more noble >than simply pandering to your `sponsors' demands. >You want us to cringe in fear upon hearing that some powerful >organizations don't approve of our convictions. >You want us to keep our place, and not protest against myopic policies >of computer vendors that try robbing our dignity. Oh please, but it gets better.... >We understand the song you sing, because we've heard it sung many >times before. >Whether it be Tiananmen Square or Panama City or Pretoria >we understand that those with a vested interest in the status quo >will stop at nothing to retain their power and control. Now Gene Spafford is the element of fascism on the net? Making empty, over melodramatic acusations can only fuel a sense of public alienation for the GNU project. I have considered in the past donating some of my projects to the GNU project, but have serious doubts about doing so now. These are largely due to the overly political tone under which software is distributed. It sure would be nice to replace those copyleft notices which add a good 8K to every file in GNU software with something like "Use this code. Improve this code. Tell me about it." Mark VandeWettering
rms@AI.MIT.EDU (06/06/89)
... But you don't need a license to do that. It's called PUBLIC DOMAIN. But, you claim, then people could snarf it up, make proprietary changes, sell it, restrict access to binaries. Yes. But you would always be able to compete with that because your intent was to make it freely available. This is small consolation for the disadvantages. If I made GCC public domain, companies would port GCC to their machines and would call the machine descriptions proprietary. As a result, only a few machines would be supported in our distribution. In other words, I would have much less leverage to persuade anyone else who starts with free GNU software to contribute to the development of free software. Even on the machines supported in our distribution, 99% of the users would get their copies from vendors who would not tell them that we exist as an alternative. I could perhaps have advanced the state of the art this way, but as for advancing the cause of **freedom to share for all the users**, it would be a failure. That would feel futile. This is why I do not develop public domain software. I am sad if, as a result, some of you don't wish to join in the project. However, I will make do. I'd rather go in the right direction with less help than the wrong direction with more help.
carl@csli.Stanford.EDU (Carl Schaefer) (06/06/89)
In article <4818@uoregon.uoregon.edu> markv@tillamook.UUCP (Mark VandeWettering) writes: > The Free Software Foundation was founded to make software > available to people everywhere, to create software that everyone > could use. Unfortunately, the FSF has decided to copyright its > works under a very RESTRICTIVE license. Hardly. There are some conditions on the redistribution of GNU and GNU-derived software; relative to software licenses from your vanilla major vendor (IBM, DEC, etc.), are you sure you can call the copyleft "very RESTRICTIVE"? > All the people who are part of the Free Software Foundation will > disagree, claiming that this license is what guarantees that the > software will remain free. But you don't need a license to do > that. It's called PUBLIC DOMAIN. See <8906022131.AA00215@sugar-bombs.ai.mit.edu> for a clear explanation of why Richard Stallman calls GNU software "free". The key point is that he is using a definition of "free" that, while somewhat different from yours, is equally valid. > But, you claim, then people could snarf it up, make proprietary > changes, sell it, restrict access to binaries. > > Yes. But you would always be able to compete with that because > your intent was to make it freely available. What if some vendor modifies a GNU product but doesn't distribute source to their changes? How does the GNU project compete with this? (see the above cited article again, re: X) The copyleft is designed to insure that all forms of GNU software remain available to all users, whether obtained directly from the Free Software Foundation or via any number of third parties. > I admire the Free Software Foundation. Their nearly endless > energy and talent has produced some very fine software, and I > probably will continue to use their products. However, I have > begun to view the "copyleft" under which gnu products are > released as being too restrictive. Who owns your kernel? (assuming you're running Unix) Do you really have more freedom to use/modify/distribute whatever OS you run than the various GNU tools you have available to you? > It sure would be nice to replace those copyleft notices which > add a good 8K to every file in GNU software with something like > > "Use this code. Improve this code. Tell me about it." This is precisely the point. The copyleft allows the first of these, encourages the second, but requires the third. >Mark VandeWettering Carl -- Carl Schaefer carl@csli.stanford.edu
plocher%sally@Sun.COM (John Plocher) (06/06/89)
+---- In <8906051911.AA00256@sugar-bombs.ai.mit.edu> rms@AI.MIT.EDU writes: | This is small consolation for the disadvantages. If I made GCC public | domain, companies would port GCC to their machines and would call the | machine descriptions proprietary. As a result, only a few machines | would be supported in our distribution. +---- In a nutshell, here is the main thrust of the GPL as I see it: The Gnu Software is available to you with source code for FREE. If you distribute the software to others, you must also make it available to others, WITH source code, for free. So far, no problems for anyone. This is what we all want. The copyleft (AKA GPL) is explicit about this - in a way that would not be possible if the software was declared Public Domain. But there is a third prong in the GPL: Any software developed using Gnu software, or incorporating parts of Gnu software, which is distributed to others must be available to them under the terms of the GPL: Free, with source, and without any restrictions on redistribution. This seems to mean that if I re-use some functions from (say) the Gnu C compiler in my pocket sized ADA(tm) interpreter, I can not make a commercial product out of it. I can live with this - after all, I couldn't take parts of AT&T's C compiler and use them in my own product either. BUT - it also seems to mean that if I use gcc to compile my software then my software is covered by the GPL. i.e., I am including Gnu header files and libraries into my code. That means that my code "IN WHOLE OR IN PART CONTAINS OR IS A DERIVATIVE OF GNU___ OR ANY PART THEREOF". (same arguments about flex, bison...) THIS IS THE REASON THAT I (as a seller of commercial software) DON'T DARE USE GNU SOFTWARE in any way relating to the products that I ship. It is because of this threat of "contamination". [1] rms's reasoning here is clear and consistent. He opposes the idea of people *selling* software. The GPL is a good way for him to enforce his beliefs. The fact that I don't agree with him completely in those beliefs does not (and can not) alter the way that the GPL is interpreted. As long as everyone understands that the GPL is designed to thwart attempts by others to comercialize the software covered by it we all can live with it. Some of us will embrace it, others will reject it. The rest of us will make informed decisions on a case by case basis. -John Plocher _________________ [1] I don't mean anything bad in the use of the word "contamination". In experimental science a sample is said to be contaminated when it is made impure by contact or mixture with anything outside the sample.
tower@AI.MIT.EDU (Leonard H. Tower Jr.) (06/06/89)
From: sally!plocher@sun.com (John Plocher)
Organization: Sun Microsystems, Mountain View
...<TEXT_DELETED>...
GPL=GNU Public License
But there is a third prong in the GPL:
Any software developed using Gnu software,
Not true. Keep on reading.
or incorporating
parts of Gnu software, which is distributed to others must be
available to them under the terms of the GPL: Free, with
source, and without any restrictions on redistribution.
true.
This seems to mean that if I re-use some functions from (say) the Gnu C
compiler in my pocket sized ADA(tm) interpreter, I can not make a
commercial product out of it.
You could sell and service it, but you would have to make ALL the
source available. You might have said: "a proprietary product out of
it".
I can live with this - after all, I couldn't take parts of AT&T's C
compiler and use them in my own product either.
great!
BUT - it also seems to mean that if I use gcc to compile my software
then my software is covered by the GPL. i.e., I am including Gnu
header files and libraries into my code. That means that my code "IN
WHOLE OR IN PART CONTAINS OR IS A DERIVATIVE OF GNU___ OR ANY PART
THEREOF". (same arguments about flex, bison...)
rms and others have straighten this out any number of times.
Particularly here on info-gcc/gnu.gcc. Probably have enough new
readers to explain this again.
Caveat: I am not a lawyer. Use your own lawyer when making legal
decisions.
When you use a program that MERELY translates one representation of a
program into another, the author of that program has nothing to say
about the legal status of your program. This means you can use
(e.g.):
- GNU Emacs to edit your own files
- gawk to do transformations using your awk scripts on your
files
- gas to assemble your files
and the GPL doesn't apply to the output. Note though the GPL does
apply to GNU Emacs, gawk and gas. That is, if you ship them or use
part of their source code with your source to form a program, the GPL
applies.
Now using Bison, lib-g++, or the (mostly un-released) GNU C Libraries
does cause the GPL to apply. You are combining your source code with
GPL'ed source code. If you don't wish to do this, you are free to
find other ways of accomplishing your task. The combined work becomes
a derived work of both your work and the work of the copyright holder
who choose to use the GPL. You both have original expression in the
combined derived work. Note that the source code being translated
into another format (e.g. ar format in C libraries) doesn't void the
GPL.
Finally, GCC. You can use GCC as long as you don't use any GPL'ed
libraries with it. GCC then is just a translator, like GAWK, GNU
Emacs, or gas.
Note we have a preference that GNU tools not be used to develop
proprietary software, but we can't legally prevent you from using GNU
tools that merely do translation. We live with it. GNU and FSF is a
legal effort to promote freed software.
A small side benefit is that such use does show off how good freed
shared software is to more people.
THIS IS THE REASON THAT I (as a seller of commercial software) DON'T
DARE USE GNU SOFTWARE in any way relating to the products that I ship.
It is because of this threat of "contamination". [1]
---------------------------------------------------------------
[1]
I don't mean anything bad in the use of the word "contamination". In
experimental science a sample is said to be contaminated when it is made
impure by contact or mixture with anything outside the sample.
Try hard to not use language that could be construed to be `bad' or
language that could be confusing. Spend a few more minutes and find a
good way of saying it. Thousands of people are reading your words.
That a fraction of a second per person!
This is much better than explaining a bad or confusing usage.
E.g. you might have said:
It is because of this threat of my software being made freed.
enjoy -len krk@cs.purdue.EDU (Kevin Kuehl) (06/06/89)
>Making empty, over melodramatic acusations can only fuel a sense >of public alienation for the GNU project. I completely agree. I love GnuEmacs, I have switched my software over to bison from yacc, etc. But, I think some of the postings may scare away those who are not as devoted to GNU as most of us have become. That definately would be a shame. The software being free is great, but the copyleft leaves questions unanswered to those of us who aren't well informed of legal practices. I won't post my questions because someone has already posted them a few articles back. (Sorry I can't remember which one right off hand.) >It sure would be nice to replace those copyleft notices which >add a good 8K to every file in GNU software with something like >"Use this code. Improve this code. Tell me about it." Another good point. Someday, hopefully this will be at the top of every source file in existance. Except Apple's :-) Kevin
fozzard@boulder.Colorado.EDU (Richard Fozzard) (06/06/89)
In article <16505@paris.ics.uci.edu> Doug Schmidt <schmidt@zola.ics.uci.edu> writes: > >...to sell-out our >morals, to prostitute our beliefs... >to prostrate ourselves before the altar of mediocrity and greed. > >...recant our faith... >...robbing our dignity... > >Whether it be Tiananmen Square or Panama City or Pretoria... >etc. Shheesh...get a grip on reality here. In no way is the issue of free software nearly as big a deal this nut makes out. People are DYING in Tiananmen Square, you know. ======================================================================== Richard Fozzard "Serendipity empowers" University of Colorado fozzard@boulder.colorado.edu (303)492-8136 or 444-3168
fozzard@boulder.Colorado.EDU (Richard Fozzard) (06/07/89)
In article <8906052357.AA00981@wheat-chex.ai.mit.edu> tower@AI.MIT.EDU (Leonard H. Tower Jr.) writes: > >When you use a program that MERELY translates one representation of a >program into another, the author of that program has nothing to say >about the legal status of your program. This means you can use >... GCC then is just a translator... This distiction of what is a translation and what is source code can get to be a very subtle one (and dont lawyers love subtleties!) Why is substituting a line of my source code with GNU-written assembly routines so legally dnd morally different from substituting a GNU-written C routine (as from a library)? ======================================================================== Richard Fozzard "Serendipity empowers" University of Colorado fozzard@boulder.colorado.edu (303)492-8136 or 444-3168
zeeff@b-tech.ann-arbor.mi.us (Jon Zeeff) (06/07/89)
>> It sure would be nice to replace those copyleft notices which >> add a good 8K to every file in GNU software with something like >> >> "Use this code. Improve this code. Tell me about it." > >This is precisely the point. The copyleft allows the first of these, >encourages the second, but requires the third. Unless you use AUX. Then the first is made difficult, the second is somewhat wasted effort and the third isn't allowed (at least on gnu mailing lists). -- Jon Zeeff zeeff@b-tech.ann-arbor.mi.us Ann Arbor, MI sharkey!b-tech!zeeff
johnson@p.cs.uiuc.edu (06/07/89)
>BUT - it also seems to mean that if I use gcc to compile my software >then my software is covered by the GPL. i.e., I am including Gnu >header files and libraries into my code. That means that my code "IN >WHOLE OR IN PART CONTAINS OR IS A DERIVATIVE OF GNU___ OR ANY PART >THEREOF". (same arguments about flex, bison...) >THIS IS THE REASON THAT I (as a seller of commercial software) DON'T >DARE USE GNU SOFTWARE in any way relating to the products that I ship. The solution is simple: write your own header files and libraries. I can't believe it would take more than a man year for a compentant programmer to recreate everything in the GNU C and C++ libraries. It would probably take less than six months. Why doesn't some enterprising small company redo them this summer and sell them to companies that are SO worried about the problem? It is true that lots of companies are worried about the problem. It is also true that there is an easy solution.
honey@mailrus.cc.umich.edu (peter honeyman) (06/08/89)
Len, let me spend a few minutes and find a good way of saying what I think about your suggestion that I use my own lawyer before deciding whether to use your freed software. I'll buy. peter