Talk:GNU/Archive 2

From Wikipedia, the free encyclopedia

This is an archive of past discussions. Do not edit the contents of this page. If you wish to start a new discussion or revive an old one, please do so on the current talk page.

← Archive 1

Archive 2

Archive 3 →

Contents 1 where to get this graphic? 2 Audio link 3 Lisp and history 4 UNIX vs. Unix 5 Merge GNU Project here? 6 See Wikipedia:Articles_for_deletion/Staff_and_employees_of_the_Free_Software_Foundation 7 Quality article 7.1 Featured Article instead? 8 GNU was founded by Richard Stallman and was the original focus of the Free Software Foundation 9 Why Hurd and why microkernel? 10 Is it complete or not? 11 pronounce? 12 UNIX popular OS? 13 Copyright

where to get this graphic?

Where can I get this image http://it.wikipedia.org/wiki/Immagine:Nuvola_apps_emacs.png on .svg format?

--Licurgo 17:21, 21 August 2006 (UTC)

Audio link

1. GNU (pronounced /gnu/ (help·info)) is a free operating system [...]
2. GNU (pronounced /gnu/; listen (help·info)) is a free operating system [...]
3. GNU (help·info) (pronounced /gnu/) is a free operating system [...]
4. GNU (help·info) is a free operating system [...] The correct pronunciation of GNU is g'noo (IPA: /gnu/), [...]

#1 is what the article currently looks like. #2 is probably what User:Aeusoes1 wanted to do. #3 and #4 are some alternatives. Which do you think would be best? --Kjoonlee 01:37, 22 August 2006 (UTC)

Another possibility is:

1. GNU (pronounced /gnu/ (help·info)) is a free operating system [...]

I think there's currently a bit of overlap between Template IPAudio and Template Audio-IPA. We only need one and whichever one it is should not underline IPA. Right now IPAudio has a place in my heart because it doesn't underline. But if Audio-IPA changes accordingly I'll be torn. AEuSoes1 02:03, 22 August 2006 (UTC)

{{Audio-IPA}} is named so because it's a counterpart to {{Audio}}. Audio-IPA now uses no underlines. --Kjoonlee 17:27, 2 September 2006 (UTC)

Why the heck is this an ogg file? Anybody whose computer automatically knows what to do with an ogg file already knows how to pronounce "GNU". I think this should be a WAV or similar... 130.15.112.215 15:25, 29 August 2007 (UTC)dan

To save space. And no they don't; people who've set up their computers that way have been mislead into thinking it's "guh-new" or "guh-noo". --Kjoonlee 17:58, 29 August 2007 (UTC)

Lisp and history

I think we need to add something to the history section. In a speech of his (but one of several sources I remember), RMS says that he initially was going for a replacement of the Lisp machine OSs, and only grudgingly decided to clone Unix:

“

About three and a half year ago it was clear to me that I should start developing a free software system. I could see two possible kinds of systems to develop: One: A LISP-machine-like system, essentially a system just like the MIT LISP machine system that had just been developed, except free, and running on general purpose hardware, not on special LISP machines. And the other possibility was a more conventional operating system, and it was clear to me that if I made a conventional operating system, I should make it compatible with UNIX, because that would make it easy for people all around to switch to it. After a little while, I concluded I should do the latter and the reason was, that I saw that you can't have something really like the LISP machine system on general purpose hardware.....So the result is then that you need something running underneath the LISP system to you catch these errors, and give the user the ability to keep on running, and debug what happened to him. Finally I concluded that if I was going to have to have a operating system at a lower level, I might as well make a good operating-system--that it was a choice between an operating system and the lisp, or just an operating system; therefore I should do the operating system first, and I should make it compatible with UNIX.

”

This would seem to be important. (Besides, does anyone else think that this idea of eventually layering a Lisp-based OS over something else might be the real motive behind the HURD?) -- Gwern (contribs) 04:08, 21 September 2006 (UTC)

UNIX vs. Unix

UNIX is written as UNIX.

See Ritchie, D. M., & Thompson, K. (1974). The UNIX time-sharing system [Electronic version]. Communications of the ACM, 17(7), 365-375.

The preceeding is available from the ACM Portal, for-pay, and may be available elsewhere. —Mike Trausch (fd0man, Talk Page) 04:58, 22 September 2006 (UTC)

Of course "UNIX" is written as "UNIX". "Orangutan" is written as "Orangutan", too - it doesn't get us anywhere. You can't use a word to define a word. Anyways, we've already been all through this a few times at Talk:Unix (notice it's "Unix", not "UNIX"):

• Talk:Unix/Archive 1#.22UNIX.22_vs._.22Unix.22
• Talk:Unix/Archive 1#UNICS_vs_UNIX

¦ Reisio 18:39, 22 September 2006 (UTC)

I do not see any actual dates on anything, and the only material I see there referenced is newer. It seems that the oldest (and most authoritative, IMHO) spelling is “UNIX”, not “Unix”. Hence, why I made the change. Note that the date on the paper is 1974. Perhaps the argument should be re-opened on the UNIX (err, Unix) page; given that this is an encyclopedia, it should use the most proper form of any noun it contains; The Jargon File and other sources certainly were written by people who were not the creators of the UNIX system, nor The Open Group. —Mike Trausch (fd0man, Talk Page) 02:07, 23 September 2006 (UTC)

It is of course your right to bring it up again, but I don't think you'll change anything. According to Ritchie and Kernighan, Kernighan coined the name. Kernighan says it was originally "Unics", a 'weak pun' on "Multics", and that eventually the 'cs' became an 'x' (a happening which no-one seems to remember specifically). As far as I know, Multics was (and is) most often formatted as "Multics" (at least when all the text wasn't all caps), which makes it seem logical to me that a pun, "Unics", would also be formed mostly of lowercase letters. It then again makes sense that any immediate corruption of this word, such as becoming "Unix", would retain the lowercase letters. On top of that you have Ritchie quoted as saying that it was "Unix" until a paper in 1974 when they typeset it as "UNIX" in smallcaps and Bell never went back. ¦ Reisio 16:17, 23 September 2006 (UTC)

Merge GNU Project here?

There is a proposal to merge GNU Project into this article. Please discuss it at Talk:GNU_Project#Merge_into_GNU. Lentower 19:44, 5 December 2006 (UTC)

See Wikipedia:Articles_for_deletion/Staff_and_employees_of_the_Free_Software_Foundation

See Wikipedia:Articles_for_deletion/Staff_and_employees_of_the_Free_Software_Foundation.

This new article started as a section in the FSF article, and was split off with no rationale, discussion, or consensus by User:Chealer (talk|contribs) .

The editors of this GNU article could have good points to make on this AfD. Lentower 18:47, 5 December 2006 (UTC)

Quality article

The verbal quality and depth of this article, as well as copyright status of images are all really good, and I feel if it were slightly better cited, it could probably get to Good Article status really easily. Any thoughts? i kan reed 18:27, 25 December 2006 (UTC)

This article needs many more in-line citations. Lentower 22:07, 25 December 2006 (UTC)

Featured Article instead?

Instead, why not get it up to Featured Article status? Lentower 22:07, 25 December 2006 (UTC)

Indeed :) A long way to go though, way too much duplication and little consistency in style (until now there wasn't even a mention of the FSF in the intro for instance). Chris Cunningham 21:00, 29 December 2006 (UTC)

GNU was founded by Richard Stallman and was the original focus of the Free Software Foundation

Thumperward:

With all due respect, your edits contain false information.

Where did you get your information? Are there other Wikipedia articles that have the wrong information?

Why don't you check out the facts, before you undo my edits? Or check out this article's History section, as my edit summary noted? Or check other Wikipeida articles, fsf.org, the WWW, etc. for the facts? Which are:

• The FSF was founded two years after Richard Stallman started the GNU Project and created the concept of the GNU operating system. Chase down the citations in the History section, etc.
• GNU was the main focus of the FSF for a while, but for more then a decade it's major focus has been furthering the cause of Free Software, of which GNU has become a very small part. The FSF is putting most of it's effort and budget into further the rights of computer uses, not producing software. Check out fsf.org, especially the press releases for the last decade, etc.

Best, Lentower 05:05, 30 December 2006 (UTC)

Apologies. (This is what I get for editing during the holiday season). My main focus was to keep the article flowing well, I wasn't paying sufficient attention to factual accuracy. The current version is markedly better. Chris Cunningham 17:05, 30 December 2006 (UTC)

No problem. I mostly wanted to get the facts right again, and make sure there weren't fixes needed to other WP articles. I haven't finished reviewing all of your changes yet. Lentower 01:20, 31 December 2006 (UTC)

Why Hurd and why microkernel?

Is it clear why HURD is the kernel of the GNU OS, and not for example Linux? Who has made the decision, when, and on what grounds? Is it a political or a technical decision? Is HURD choosen because it is a microkernel system? Aren't there other microkernel systems which are more functional than HURD, and that are licenced under the GPL? I think it would be good if the decision making could be clarified on either the GNU page or on the GNU Project page. --HelgeStenstrom 12:23, 2 January 2007 (UTC)

Richard Stallman made the decision. The reason is that they were working on the HURD before development of Linux started, so Richard decided not to cancel the HURD project. A microkernel design was chosen because it looked like it would be easier to debug (due to being mostly user-space) and because Richard found half of a microkernel that GNU could use (mach) so it looked like half the job was done already. I think these questions are answered in greater length in the GNU/Linux FAQ. Gronky 12:34, 2 January 2007 (UTC)

Some issues that are worth mention besides "Stallman said so", include:

• A wholly unique and separately written kernel avoids being encumbured by copyright and patent enforcement that occurred with early free versions of BSD[1] and is now occurring with Linux.
• Besides debugging, there are a whole host of technical features that make microkernels worthwhile.[2]* A microkernel design, as Mach and the Hurd provide it, give greater freedom to users of the system by allowing them to do more system level tasks in normal user space, like mounting their own file systems or running their own instances of servers. As Marcus Brinkmann noted in a GNU Hurd talk "But even on [most systems] consisting only of free software, you can not easily use the filesystem format, network protocol or binary format you want without special privileges. In traditional unix systems, user freedom is severly restricted by the system administrator."[3]

These could be incorporated here or at GNU Hurd. --75.69.87.211 16:40, 2 January 2007 (UTC)

Since they made the decision 20 years ago, I'd say they simply had no other choice than writing their own. I actually recall seeing it somewhere, so I searched around and found this: The Hurd and Linux. 80.233.255.7

When they saw hurd failing and linux taking off they could have accepted linux as the official kernel for gnu, it seems they didn't mainly for political and philosophical reasons. Plugwash 00:46, 13 January 2007 (UTC)

Linux was an unportable, very i386-oriented kernel at the time, so initially it might have been a technical decision. But politics and philosophy probably played their parts as well. Can't blame them, though. Without these two they could've just kept using proprietary software. 80.233.255.7 02:04, 13 January 2007 (UTC)

To say this another way, wthout the politics and the philosophy, there be little FOSS software. No Linux, No GCC, No FOSS versions of BSD and Unix. No ................ Freedom comes with it's own costs. Lentower 06:24, 13 January 2007 (UTC)

This is sort of going off on a tangent but I don't really think that's completely true. FOSS has practical benefits that go beyond abstract notions of freedom. —The preceding unsigned comment was added by 152.30.221.68 (talk) 02:02, 3 April 2007 (UTC).

Is it complete or not?

Hi, I was just passing through and I noticed that the opening paragraph of this article says that "GNU ... is a complete computer operating system", but then in paragraph 3 it says "a complete GNU system has not yet been released. The official kernel is the GNU Hurd. However, Hurd is not yet finished". There seems to be some sort of contradiction here. -- Sakurambo 桜ん坊 00:09, 17 January 2007 (UTC)

Yes that looked odd to me, too, but in a simpler way: 'complete' is superfluous and/or irrelevant. ¦ Reisio

I'll go ahead and fix that then. -- Sakurambo 桜ん坊 11:42, 17 January 2007 (UTC)

My bad on this one I think. It's important to note that completeness is a very important factor in the original project goals, although now that a completely free software OS is a possibility without having to go to GNU for everything this has gotten somewhat muddy. Chris Cunningham 12:06, 17 January 2007 (UTC)

The GNU project's mission of making a complete free software OS exist was completed in 1992. It was originally thought that making a complete OS exist would entail writing a complete OS, so it was expected that the OS and the mission were pretty much the same thing and would be completed together. When the mission was completed by a chance third-party project in 1992, the GNU OS was still not completed. I'll check the article to see if anything needs fixing, but I wanted to explain this here first. Gronky 16:11, 17 January 2007 (UTC)

That makes complete sense as a mission statement, but there where it was said sounded awkward even if it were true. -- AdrianTM 16:25, 17 January 2007 (UTC)

pronounce?

Guh-New or New? —The preceding unsigned comment was added by 75.15.251.155 (talk) 22:02, 21 January 2007 (UTC).

Neither. It's hard g followed by "noo". -- parasti ^(talk) 00:05, 22 January 2007 (UTC)

I added text based pronunciation (guh-noo), if you plan to remove it please consider that not all people can hear the verbal pronunciation. 75.15.234.217 16:59, 28 January 2007 (UTC)

GNU is pronounced that there is only one syllable. -- mms 17:39, 28 January 2007 (UTC)

"GNU is a recursive acronym for “GNU's Not UNIX”; it is pronounced guh-noo, like canoe." [4] 75.15.234.217 22:17, 28 January 2007 (UTC)

FWIW, here's a link where RMS describes how GNU is to be pronounced: http://fsfeurope.org/documents/rms-fs-2006-03-09.en.html#the-name-gnu (just after the 49:21 timestamp). He says it's not like "noo". Gronky 03:28, 29 January 2007 (UTC)

Please visit Google Videos and watch any speech given by Richard Stallman. (There are a couple of short ones.) Regardless of what any written document might say, it is in fact one syllable as mms above points out. (By the way, I'm not trying to suggest that it be pronounced "noo"; I don't think anyone is.) 80.233.255.7 03:41, 29 January 2007 (UTC)

Strange. I've listened to tens of speeches by him. Maybe we're using different descriptions of the same thing. I'd describe his pronunciation as being two-syllable, and being like "canoe". Do you regard "canoe" as being single syllable? Or are you saying that his pronunciation of GNU is not similar to "canoe"? Gronky 04:11, 29 January 2007 (UTC)

No and yes. Unlike "g" in "gnu", a vowel follows after "c" in "canoe", hence two syllables. For comparison, look at a simple way to pronounce "lambda" that is sometimes suggested. It's "lam-duh". "Duh" here stands for "da". From there, "guh-noo" suggests that "g" in "gnu" be pronounced in a similar way, but this is not the case, according to how Richard Stallman himself pronounces it. 80.233.255.7 16:07, 29 January 2007 (UTC)

Hmm. So we here him differently. When I hear him saying "GNU", it sounds like "canoe" (in terms of syllables). BTW, whether a vowel follows the g or the c is irrelevent. If the spelling of "canoe" was changed to "cnoe", the pronunciation or number of syllables wouldn't necessarily change. Gronky 16:13, 29 January 2007 (UTC)

The difference is completely irrelevant in English anyway. "cnoo" and "canoe" would be phononyms in most accents. Chris Cunningham 16:18, 29 January 2007 (UTC)

I understand that it's irrelevant, I was just describing the logic behind my reasoning. In any way, "gnu" does not sound like "canoe" to me. 80.233.255.7 16:34, 29 January 2007 (UTC)

Nor it does to me, besides I would say canoo and cnoo differently, but English is not my first language... guh-noo strikes me as incorrect though from my rudimentary knowledge of phonetics. -- AdrianTM 18:32, 29 January 2007 (UTC)

I'm not a native speaker either. In fact, I think that's exactly why the difference in pronunciations of "canoe" and "gnu" seems obvious to me. 80.233.255.7 14:49, 30 January 2007 (UTC)

RMS says it's /gnu/. Link: Talk:GNU/Archive 1#Pronunciation --Kjoonlee 04:39, 29 January 2007 (UTC)

GNU doesn't sound like canoe (I always pronounced as one syllable like "new"), but if the official website suggests it then thats how they want it to pronounced, you say pah-tate-toe, i say poh-tate-toe :P —The preceding unsigned comment was added by Ismellbeans (talk • contribs) 01:53, 30 January 2007 (UTC).

I'm one of the people who have been deleting mentions of "guh-noo," based on RMS's comments. My rationale is thus.

• Arguments for IPA /gnu/:
  • Ad-hoc pronunciation guides are generally problematic, because there's always a chance of confusion ("/gəˈnju/")
  • RMS favours /gnu/, not /gəˈnu/

However, if I think hard enough I can't can think of several arguments against it.

• Arguments against IPA /gnu/:
  • The WP:NOR, WP:REF, and WP:V policies rule against including comments from message boards or talk pages, and we can't include info that hasn't appeared somewhere else before
  • Published sources mention guh-noo (*used to mention "new with a hard g" or "guh-new")

What should we do? I'm tempted to invoke WP:IAR, but I'm afraid that might not be very nice. --Kjoonlee 11:17, 30 January 2007 (UTC)

Why don't we split the difference? The text should contain both versions, indicating that gnu.org says "it is pronounced guh-noo" and Stallman (according to others here) prefers "noo" --Nshewmaker 14:21, 11 April 2007 (UTC)

The problem is, we can't. :( If we were to show both versions, we will have to show a source for /gnu/, but we won't be able to provide reliable sources, so we can only mention /gnu/... --Kjoonlee 17:26, 6 September 2007 (UTC)

UNIX popular OS?

"At the time, Unix was (and is) a popular operating system"

• First of all I don't know what you mean by UNIX being popular, Solaris and BSD are not considered quite "popular", and I'm not sure if Mac can be considered pure "UNIX" (and it's debatable if it can be considered "popular" either), to me this sounds like POV pushing.

• Second issue, I don't like the "At the time, UNIX was (and is)", if you decide that's true it's a popular OS then use "UNIX is a popular OS" not "at the time was (and is)" which sounds unnecessarily complicated. -- AdrianTM 19:49, 11 February 2007 (UTC)

Alter the wording if you wish. Flavours of UNIX are still probably the most popular proprietary OSs after Microsoft. Running on enterprise servers, UNIX holds a considerable share of the market. The wording before I made the change suggests the UNIX has suffered the same fate as CP/M. --Philip Baird Shearer 20:42, 11 February 2007 (UTC)

BTW What makes you think that UNIX was more popular in 1983 than it is now? I suspect that if the article is going to designate the OS as popular in 1983, then it is still popular now because there are, I suspect, far more copies of UNIX running today than there were in 1983. --Philip Baird Shearer 20:52, 11 February 2007 (UTC)

I use Opera, there are other couple of millions of Opera users probably more than the no. of users 5 years ago, however that doesn't make it "popular". I didn't find recent numbers but in 2000 it looks like 15% of servers were UNIX servers. I would only assume that the procentage was bigger in the past since they didn't have competition from Windows and Linux server (at least not in 1983). I think it makes sense to say "at the time, was a popular OS" because it was more popular than now. However, this is not a big issue, I'll let other people see the discussion and decide how this small issue should be decided. -- AdrianTM 00:50, 12 February 2007 (UTC)

Definition time. Would "Unix-like" be a useful term? In that case, unix-like operating systems are quite popular at this point in time. Especially counting embedded systems. --Kim Bruning 12:13, 14 February 2007 (UTC)

Trouble is that GNU is Unix-like and this could become a circular argument it is after all GNU not UNIX. It really depends if one believes UNIX to be a set of copyrighted code or an implementation of a methodology. Personally I'd go with the second. all lowercase because the shift key takes more effort, short names to save typing (very useful on very slow connections), multi user very useful when playing games, or in the words of the the early 80's email rpdup "Unix is a lot more complicated of course -- the typical Unix hacker never can remember what the PRINT command is called this week -- but when it gets right down to it, Unix is a glorified video game." --Philip Baird Shearer 17:53, 14 February 2007 (UTC)

I think some people conspired to redefine unix, since the popularity of gnu/linux. ;-) --Kim Bruning 18:25, 14 February 2007 (UTC)

GNU is a POSIX system, not a UNIX system. Noahslater 12:53, 15 November 2007 (UTC)

Copyright

"The FSF also holds the copyrights for some GNU software packages."

I think this is worth clearing up. The FSF requires either signed copyright assignments or signed copyright waivers for contributions to many important packages. Which packages? As far as I can tell, there are three categories of GNU software: software which the FSF owns, third-party software which is officially sanctioned (the lead says X.org and TeX), and third-party software which is not (Linux kernel). It would be good to clarify exactly which packages. Furthermore, is there a fourth category: official GNU software that the FSF does not own, but that is not third-party software either? To me, this sentence implies that it might exist. If so, what is it? Thanks, Chris Pickett 21:10, 18 February 2007 (UTC)

You're basically right. FSF requires signed copyright assignment papers from contributors to GNU packages. And if you think your employer or school might have some ownership claim to code that you right, FSF asks you to get your employer or school to sign a waiver of ownership claims with regard to whichever specific GNU project you're contributing to. So FSF owns the copyright for GNU software, but there are a handful of exceptions - I only know of two. One is the MULE component of GNU Emacs. This was written by the Japanese government, and they can't assign copyright, so the code was added to Emacs without assignment[5]. The second is GNOME. GNU launched GNOME in 1998, but hundreds of people contributed without assigning copyright. Getting GNOME working was important, and the wave of developers were doing good work, so FSF didn't interfere. AFAIK, GNOME is the only big project that GNU launched but for which it doesn't hold the majority of the copyright (although they hold the copyright for some parts, and other parts say "copyright FSF" even thought no copyright assignment form was actually signed). So "some" is misleading as it implies a much smaller amount than FSF actually hold. It would be more accurate to say: "FSF holds the copyright for almost all GNU software".

Then there's the related but separate issue of what is and what isn't an official part of the GNU OS. GNU OS is a collection of existing software and software written by the GNU project. X.org and TeX are two large existing pieces of software that GNU incorporated. Linux kernel is not part of GNU at all, but the main way that people use GNU is in combination with the Linux kernel - so most GNU users use a Linux-based variant of FSF's GNU OS. Gronky 01:59, 19 February 2007 (UTC)

So how about: "The FSF holds the copyright for all GNU software. Contributors to GNU software must sign copyright assignment forms, which employers and educational institutions must sign as well. The only exceptions are the GNOME project and the MULE component of GNU Emacs. The FSF claims it requires this assignment to defend itself properly in the event of a lawsuit, and to make strategic licensing decisions." I think that clears up how and why the FSF holds all the copyright. If there are other exceptions, someone will fix this sentence; if it says "almost all", it's still kind of vague. (Also, GNOME is a huge undertaking, so that needs to be mentioned.)

btw, I see you've basically done this part now, but maybe the motivation for the assignment should be explained? Chris Pickett 19:46, 19 February 2007 (UTC)

Ok, I've done this now. I've also moved the bit about copyright assignment from the History section into the section about licences, and have made that a section about "Copyright, licences, and stewardship". It looks like all that information should go together. Gronky 03:19, 21 February 2007 (UTC)

As for the existing/written-by issue, I think explicitly listing the pieces of existing software that the GNU project did not write that are officially part of the GNU OS would be useful; if we can't find this information anywhere then I think the text needs clarification somehow; if it just includes all software components that GNU is not directly responsible for, it should say that instead. (Note that since GNU is directly responsible for the kernel, this excludes Linux.) Cheers, Chris Pickett 19:32, 19 February 2007 (UTC)

In the development section I've made two subsections, one about the third-party components, and one about Linux, Hurd, and the role of kernels. Gronky 03:19, 21 February 2007 (UTC)

By the way, the FSF doesn't "own" software. From Gronky's response, "holds copyright" is what you really mean. 80.233.255.7 01:02, 20 February 2007 (UTC)

Holding copyright is generally equated with "ownership" of that aspect of the "intellectual property", but sure, "holds copyright" is a clearer term and I also prefer it: "own" has serious potential to confuse, witness, "I own a copy of Windows XP!" --- no you don't! Chris Pickett 08:43, 20 February 2007 (UTC)

I think "own" is the correct term in this case, for example Microsoft owns Windows XP and grants use rights to people who buy a license. You don't own Windows XP, you own a usage license and the rights that license grants you (nothing more, nothing less) -- AdrianTM 14:21, 20 February 2007 (UTC)

I don't think people can "own" licenses. I think people own *copies* of Microsoft Windows, just like you own a copy of a book. However, Microsoft and other propprietary software companies have a copyright and license that is a lot more restrictive (and sold as a commodity), thus the reason you're inclined to consider the license the thing you own, but it's really not. Bulk licensing for example gives one the right to own (and use) multiple copies of the software.

In conclusion, if Microsoft owns anything, it owns the temporary rights granted by copyright, patents and trademark law on its software. And occasionally they relinquish those rights for some of their binary-only libraries or source header files. --69.54.29.23 14:32, 20 February 2007 (UTC)

Well, anyway, maybe I should have dropped the "a copy" part but it seems that definitely "holds copyright" is less ambiguous to everybody. You "own" the physical medium on which the copy is reproduced, same goes for a book, but nothing more. You don't own a license, that's just a contract you've agreed to. The copyright is intellectual property (much as that term is bad in its own right) and it can be "owned" independently of physical property. So I think it's fair to say, "I own this piece of code," but it's more clear to say, "I hold the copyright over this piece of code." (In fact, what I actually think is that only source code should be copyrightable, and copyrighting binaries is like copyrighting a house, but that wouldn't stand so well with the establishment.) Chris Pickett 19:33, 20 February 2007 (UTC)

"The GNU Project requires that contributors assign the copyright for GNU packages to the Free Software Foundation,[9] although exceptions have been made in the case of MULE,[10] and large parts of GNOME. Most GNU packages are licensed under the GPL, while a few use the LGPL, and an even smaller amount use other free software licenses." This seems to be in conflict with information on the GNU website: "For a program to be GNU software does not require transferring copyright to the FSF; that is a separate question. If you transfer the copyright to the FSF, the FSF will enforce the GPL for the program if someone violates it; if you keep the copyright, enforcement will be up to you." The reference cited for the claim in the article, I believe is referring not to GNU as a whole, but to Emacs in particular (which does require assignment usually, but an exception was made for MULE.) Are there objections to me tidying the claim in the article, so that it states that some existing projects require assignment, but that GNU in general does not? 82.32.8.187 22:56, 8 April 2007 (UTC)

There are two issues you are discussing. The article is trying to explain the issue of assignment of contributions to existing GNU packages. You're mentioning that copyright assignment is not required by the FSF acceptance as a GNU package. My experience is that most GNU packages require copyright assignment, therefore the article is accurate. You could mention in the article that "new packages do not have to accept copyright assignment to become GNU packages. This will forego copyright enforcement by the FSF." --75.69.87.211 18:11, 9 April 2007 (UTC)

In my experience, most GNU packages use C, so I'll make a change to the article to say that "The GNU project requires the use of C". (Obviously, I WON'T really make this change, I'm trying to demonstrate the flaw in what you are saying :) ). The article is not accurate, because it says "The GNU project requires that", about something the GNU Project does not require. It may be the case that most GNU packages require it (just like it is the case that most use C), and I'd be happy if the sentence said "Most GNU packages require copyright assignment...". 82.32.8.187 07:00, 13 April 2007 (UTC)

I have change the wording of this section. Neither the FSF or the GNU Project require copyright assignment for GNU software. The only time copyright assignment is required is when a new contributer joins an existing project for which the FSF owns copyright.