Path: utzoo!utgpu!jarvis.csri.toronto.edu!rutgers!gatech!tut.cis.ohio-state.edu!pacific.mps.ohio-state.edu!gem.mps.ohio-state.edu!ginosko!uunet!mcsun!sunic!tut!santra!sja From: sja@sirius.hut.fi (Sakari Jalovaara) Newsgroups: gnu.misc.discuss Subject: Re: Why I do not support GNU Message-ID: Date: 19 Oct 89 11:03:11 GMT References: <8910160520.AA01740@sugar-bombs.ai.mit.edu> <1989Oct18.080236.23848@rpi.edu> Sender: news@santra.UUCP Distribution: gnu Organization: Helsinki University of Technology Lines: 157 In-reply-to: tale@pawl.rpi.edu's message of 18 Oct 89 08:02:36 GMT In article <1989Oct18.080236.23848@rpi.edu> tale@pawl.rpi.edu (David C Lawrence) writes: > SJA> So it *is* OK to disagree with FSF in gnu.misc.discuss after all. Wow. > > It has been right from day one; the FSF has not ever claimed otherwise. Actually, it hasn't and they have; not on day one, but it seems this policy has changed. Wow. Here are a couple of quotes from article <330@ncis.tis.llnl.gov> posted to news.admin and gnu.misc.discuss: > [From Usenet document "Alternative Newsgroup Hierarchies":] > > operating system with freely redistributable source code. The GNU > > ! Project is led by Richard Stallman. Note that use of these groups to > > ! discuss topics considered contrary to GNU aims and political > > ! philosophy are considered off-limits (e.g., porting of GNU code to > > ! Apple machines, usefulness of intellectual property laws). The writer of the article says: > I am informed that the addtional sentence was written by > either Len Tower or Richard Stallman. I have the entire article if you want it; it may also be in the gnu.misc.discuss archives (possibly anonymous ftp in tut.cis.ohio-state.edu:pub/gnu/misc-discuss/gnu-misc-discuss.) ANH now says it is OK to disagree with FSF in gnu.misc.discuss, not in other gnu groups. > He will do what he thinks is right in the pursuit of his goals. No > one forces people to use FSF code. If the users find the terms > unacceptable, they will not be users any longer. It is their right. Exactly. "Why I do not support GNU" listed some of the reasons why *I* find their terms unacceptable. > SJA> Again, the same person proposes to block messages from anyone working > SJA> for Apple to the newsgroup gnu.gcc: > > You (or perhaps I, but I think the former) have misunderstood the > [now defunct?] proposal to filter traffic on the GNU lists. It was > not to block messages from people working at Apple, but to block > messages which would counter the FSF's position regarding Apple. I have article <8905310246.AA00550@sugar-bombs.ai.mit.edu> in gnu.gcc; written by a member of FSF, proposing the blockade. It does say > I am considering blocking messages which actually try > to block our work, such as his previous message. but still it leaves me the impression that Apple is considered the Evil Empire and "Apple wants to say" and "Apple can do." I do not know whether the block was supposed to work automatically by blocking every message from the domain "apple.com" or whether the group was supposed to be monitored (by a supporter of FSF, I would presume) and only "harmful" messages blocked. The FSF media policy that FSF and only FSF is allowed and expected to make political "announcements" in gnu.emacs still strikes me as odd. Others seem to find it good and proper; this is certainly their right. > If the source > code _necessitates_ compiling or linking with FSF code then it is the > student's own problem for writing it that way. I remember seeing an article saying that if a bison input file uses bison-specific features the input file falls under the GNU copyright. So FSF claims they have created *proprietary* programming interfaces? They claim intellectual property rights on the "%expect" keyword? Ditto for G++ libraries? Does anyone have the statement from FSF that says this, it seems I didn't save it. > If the terms of the > FSF were unacceptable then it should have been done differently. Quite; the point is that if a programming assignment requires the use of a tool, a non-GNU tool must be provided (or, if I understand the FSF claim to the "%expect" keyword correctly, an equivalent tool must *exist* so that FSF cannot claim ownership of the code.) > Regardless, it would not be illegal for the instructor to assign a > project which required linking with G++ libraries. Intellectual right to a library interface? The student must be careful not to use libg++ -specific constructs. Hopefully these are clearly documented. I think libg++ is a sad case; not only does it make a compiler less useful for some of us, it makes a whole language unavailable to many. > SJA> You can do anything with public domain code, including sell it. > > Strangely enough, you can sell Richard's code too. I wasn't clear enough on the "GNU copyright confusion" bit. What I wanted to say was that if I write code derived from GNU code I let *someone* *else* to decide how *my* work (the parts I wrote, not the parts someone else wrote) are to be distributed. Say I don't want my code to be sold and FSF wants to let anyone do it. Say I want to prevent (if possible) the user-does-the-link trick and FSF doesn't. Say I want to write completely free (public domain) code and FSF doesn't. Clause 2b in the GNU copyright (or should I say license?) says that FSF wants to control *my* work. I wrote (beginning of) a FORTRAN front-end for GCC (oh no! FORTRAN! GAG! Well, FORTRAN isn't *that* bad (oh yes it is, take it from me: I wrote a damn parser for it!)) I'm not giving it away or continuing writing it because I can't make my part of the work public domain. Not their part, *my* part. The GNU Manifesto talks about control over other people's lives being a Bad Thing. Weird. > SJA> * The modifications are so complex you can't duplicate them. > SJA> In this case, the users can choose between your free but > SJA> not-so-great program and the New and Improved commercial > SJA> version. Without competition, they only would have one choice > SJA> - the incomplete version. The more competitors there are, the > SJA> better (and cheaper, etc) programs they have to make to > SJA> stay in the competition. > > I disagree with the conclusion you draw in the middle of this > paragraph. It is not necessarily at all the case that without > competition they would have only one choice. There is a lot of > excellent free software development which goes on which has absolutely > nothing to do with competition. Certainly, there may be several different free versions. But the commercial version is *one* *more*. Nothing wrong having one more choice? I would (perhaps incorrectly) expect that most often a company does not make a decision between commercial/free but between commercial/don't write it at all. Some people would make the decision to go commercial with a public domain gcc but, because gcc is GNU copyrighted, now release it under the GNU copyright instead. This is a trade-off; it just so happens that I think the advantages of public domain are great enough to justify suffering the Greedy Hoarder's commercial version. Question: FSF probably has a list of porters of gcc. How many of those do you realistically think would have made their version commercial, were it possible? How many of those versions (different back-ends) would have been re-made and released as public domain, giving users a choice? Are so many of the porters greedy hoarders? +++ It seems to me that some people are saying they would like to contribute work to a free operating system but, for some reason, don't contribute to GNU. Doesn't that strike you as odd? Doesn't that suggest that just maybe there is something wrong with GNU? ++sja