Talk:GNU/Archive 3

Archive 1

Archive 2

Archive 3

Archive 4

Incomplete

I don't think it's appropriate to say that GNU is "incomplete". Only reason to say that, that I can think of, would be Hurd. Yes, Hurd is far from usable, but it works pretty good on a basic level. I've booted and messed around with Hurd Live CDs myself. I'll remove the statement. 80.233.255.7 21:36, 20 February 2007 (UTC)

No, wait. I guess I was viewing a cached version. "Complete but unfinished" looks better. Oops. Ignore away.  :) 80.233.255.7 21:39, 20 February 2007 (UTC)

Ignored.  :) Chris Pickett 22:14, 20 February 2007 (UTC)

User Yworo is well know for pushing POV and then accusing other editors of it. The use of the word "incomplete" is not appropriate in the definition of the GNU operating system. Should we also add "incomplete" to the definition of "Linux" as an OS since indeed Linux is only a kernel that by itself is not a complete OS??? Using the opinion of a third person that is in accordance with his own POV is not acceptable as a valid and objective source.

--Grandscribe (talk) 09:57, 30 June 2010 (UTC)

Readers will only be helped if the meaningful outlooks on this are cited and echoed in the text in a neutral way. The article (lead) does not yet do this. Some of it has to do with definitions. Linux, for example, is not an OS, it's a family of OS distros which share the Linux kernel. Calling Linux "incomplete," while in some ways true, would be wholly misleading. Likewise GNU. In a way, yes, GNU is incomplete because the kernel isn't ready, but it would be misleading to make a sweeping assertion in the article's narrative voice that GNU "is an incomplete OS." Lastly, GNU is still most widely noted as a port to other OSs, for its apps, utils, interface and so on. All of this should be pithily dealt with by the text. Gwen Gale (talk) 10:35, 30 June 2010 (UTC)

After reading this, I would settle for "unfinished" rather than "incomplete" in the lead sentence. Once there is an official production release of GNU Hurd, it could of course be removed. Yworo (talk) 16:09, 30 June 2010 (UTC)

Every active open source OS is "unfinished" technically, being always in development and being new features/drivers/bug fixes always added. Completeness is not a meaningful property therefore. GNU/Hurd boots, runs and can run GNU utilities just fine; it is simply hard to use, unstable and has only disadvantages compared to the Linux or *BSD kernels. Therefore, usually the distinction is between being production ready or not; I'd say the best wording is "not yet production ready".--Cyclopia^talk 10:21, 1 July 2010 (UTC)

The first sentence needs to deal with this. I still think "unfinished" is best. But would consider "as yet unreleased for production". However, simplicity is always best in the lead sentence, so "unfinished" followed by an explanation in the first paragraph of how it is unfinished would be better than any long complicated phrase in the first sentence. Yworo (talk) 12:52, 1 July 2010 (UTC)

Moved from User talk:Cyclopia

Hi. Perhaps you are unaware that GNU will not run without some other kernel, such as Linux? An operating system, by definition, consists of a kernel plus utilities. Since GNU Hurd is not yet functional and probably never will be, GNU is either something other than an operating system, or it is not yet complete. Even Stallman will admit this if pressed. It's why he is so obsessed with calling Linux "GNU/Linux". Yworo (talk) 13:48, 30 June 2010 (UTC)

GNU Hurd is functional, I installed it myself several times and it is definitely functioning. It is not very stable and it is abysmal when dealing with drivers, but you can run programs on it (heck, even KDE and GNOME can run on it!) and use it for simple everyday tasks. That said, "completeness" is POV : one could argue that Linux too is incomplete, or that Hurd is complete. To define it incomplete, you should have an objectively defined set of features that are required to be called so. Since I am not aware of an official and/or objective definition of operating system completeness apart from being able to be compiled and boot (which Hurd does), I'd say we should avoid any "complete/incomplete" characterization as POV and/or misleading. --Cyclopia^talk 09:54, 1 July 2010 (UTC)

Unfinished and yet to release a first production version, then? Yworo (talk) 12:41, 1 July 2010 (UTC)

Remove "unfinished" (which is synonimous to "incomplete" and as such inappropriate), and I'm fine. --Cyclopia^talk 15:38, 1 July 2010 (UTC)

I'm happy with Gwen Gale's edit, "is a computer operating system composed entirely of free software, with an as yet unfinished kernel." I think it's entirely accurate. Yworo (talk) 15:42, 1 July 2010 (UTC)

It isn't. Or better, it is misleading: Linux is no more "finished" than GNU Hurd, in the meaning that both are in ongoing development, even if at very different stages. --Cyclopia^talk 15:43, 1 July 2010 (UTC)

Sorry, have to disagree. GNU Hurd has had no official production release. It's underdevelopment and thus "unfinished". Linux is under continuing development, yes, but had an official production release years ago. Unfinished is exactly the right word for it, just a a building can be complete (i.e. structurally) but still be unfinished (cosmetically). Except that Hurd isn't even that far along. Yworo (talk) 15:56, 1 July 2010 (UTC)

Find an objective definition of "unfinished" applied to operating systems and we can agree. Until that point, your disagreement is only POV. To be honest: If we were chatting at a beer, I would not mind or I would very much agree with you, because colloquially it makes much sense what you say. But since we're writing a neutral encyclopedia article, I must worry about misleading and personal statements. Hurd can be used, hundreds of useful programs run on it and it can be installed on real computers, but it is defective in modern drivers and some other goodies: as such it is no more "unfinished" than, say, MS DOS. --Cyclopia^talk 20:38, 1 July 2010 (UTC)

"Unstable kernel"? That can be sourced. Gwen Gale (talk) 22:52, 1 July 2010 (UTC)

User Yworo is not trying to reach consensus nor acting in good faith. He insists in pushing his POV to include the word "incomplete" even though most editors strongly disagree. As sources he gives links to articles which are opinion of the authors that coincide with his own personal opinion and not undisputable facts. --Grandscribe (talk) 12:45, 5 July 2010 (UTC)

I would agree there is no consensus for incomplete in the lead, but sourcing it as a published opinion in the body of the text would likely be ok, if other opinions are also sourced there. Gwen Gale (talk) 12:48, 5 July 2010 (UTC)

It's a current element of what GNU is, and really belongs in the lead paragraph. No one has brought forward sources which say that it is "complete". Why are we excluding a word without sources for the opposing opinion? I have no doubt that we can found sources that say that the goal of GNU is a complete operating system, but where are the sources that say that it is now a complete operating system? As for working toward consensus, several editors have been involved in trying to fine tune the wording. I find that the last edit by Lawrencekhoo was much better than what I started with. Yworo (talk) 15:09, 5 July 2010 (UTC)

You ask for a logical fallacy. What you ask is akin to "Do we have sources which say that Obama is not a green-eyed butterfly?" and taking the obvious lack of these sources as a proof that we can state that Obama is a green-eyed butterfly. Completeness and incompleteness are not adjectives that have an objective, non-OR, NPOV definition applied to operating systems. If it is incomplete, can you point us 1)what is it currently lacking and 2)why are these lacking features crucial to be "complete"?

Because for example our operating system article says: "An operating system (OS) is a set of system software programs in a computer that regulate the ways application software programs use the computer hardware and the ways that users control the computer." - GNU does all of that. Also it has all components listed in Operating_system#Components. So in what meaning is it incomplete?

About the source: To quote the source in context, something like "it has been labled as 'incomplete' by some[ref]" is OK to me. To state unequivocally that it is incomplete, like not meeting an objective definition of completeness, is totally misleading. --Cyclopia^talk 15:28, 5 July 2010 (UTC)

What you are saying is that we should favor your original research and synthesis over what multiple sources actually say? Gwen Gale said in another thread that the wording in the source is what should be used. Please by all means provide opposing sources, ones that say GNU Hurd is complete and completes GNU. Sources that are not affiliated with the GNU project or the FSF. Yworo (talk) 15:42, 5 July 2010 (UTC)

This is a strawman argument. I am advocating no OR and no SYN, quite the opposite: to say that it is "incomplete" is OR because it relies on a yet-to-know definiton of "completeness" that you are (not surprisingly) refusing to explain. Now, your source is quite appropriately entitled "Opinion: The top 10 operating system stinkers". It is an opinion piece, therefore, obviously, not a technical NPOV source. Therefore it is perfectly OK to use "the wording used in the source" to say something on the lines of "It has been considered "incomplete" by commentators[ref]", but it is emphatically not suitable to source the statement that it is "incomplete" like it was a fact. --Cyclopia^talk 16:27, 5 July 2010 (UTC)

(moved from below): [1]: The current state is still far from being on par with Debian's established Linux ports, but it is mostly up to date and reasonably usable. - It doesn't say it is "complete" because (again) completeness and incompleteness have no objective definition applied to OS, but it shows that a GNU OS with a GNU kernel exists and it is (somehow) usable. --Cyclopia^talk 16:41, 5 July 2010 (UTC)

It's a project status update, not an objective analysis or educated opinion. Yworo (talk) 20:08, 5 July 2010 (UTC)

Operating System?

How many people consider it to be an 'operating system'? It is POV to simply assert that it is one. To many, it's just considered to be a series of userland programs. -Nathan J. Yoder 19:04, 26 March 2007 (UTC)

Do you know what an Operating System is? -- AdrianTM 02:24, 27 March 2007 (UTC)

Errr, what? The ultimate goal is a complete operating system. Don't let the FSF's insistence on conflating Linux and GNU distract you from the fact that their ostensible goal is to do the whole thing themselves. Your assertion is sort of like saying a lot of people think of BSD as a collection of userland programs because almost everyone uses bits of BSD code in other operating systems. Chris Cunningham 12:54, 29 March 2007 (UTC)

Small correction: FSF and GNU never aimed to write the whole thing themself. They just wanted to do what was necessary to make a free OS exist. This included using X Window System and TeX and some other software they didn't write. Gronky 13:08, 29 March 2007 (UTC)

I doubt TeX can be considered as a part of an Operating System, it's only an application. -- AdrianTM 13:36, 29 March 2007 (UTC)

Errr, please don't let's get into this again. Thanks :) Chris Cunningham 13:42, 29 March 2007 (UTC)

It doesn't really matter what your personal view is, as this is about NPOV policy, which represents views as they are and it's not universally accepted that GNU is an operating system. Out of curiosity, who thinks of BSD as a collection of userland programs? -Nathan J. Yoder 23:49, 29 March 2007 (UTC)

It was an analogy. Anyway, feel free to find some actual sources who don't, err, believe that GNU is an operating system and we can talk. This is a highly silly argument from where I'm sitting. Chris Cunningham 10:17, 30 March 2007 (UTC)

I did a quick search and found this (as an example) right away. This is entrenched with the whole debate over Linux vs. GNU/Linux and whether Linux is an OS or not. I'm very surprised that you've never come across this. I don't find it at all silly to consider userland programs to not be an OS. It's actually rare that I hear anyone refer to GNU as an OS, which is why I was surprised to see it say that on the Wikipedia article.

I should add that if you look at that particular thread, there are several people agreeing with him as well. You can weed through the larger of that and other threads if you wish, but it's rather tedious--you have to read all the other stuff that comes with those arguments as well. If you're curious, I searched for gnu "is not" operating system. You can try varying searches. -Nathan J. Yoder 07:18, 1 April 2007 (UTC)

You're confusing issues here. If you use Linux as a kernel then there's some dispute as to whether the whole thing is really "GNU" (Gronky thinks it is, I don't really care but I'm not willing to call it that, etc). But if you're using an officially-approved GNU kernel (like Hurd) from the FSF (and leaving out silly debates about X and tex for the sake of brevity) there's no question that the combined result is "an operating system called GNU". See the voluminous archives on talk:Linux for some discussion of the issue you're on about. Chris Cunningham 09:23, 1 April 2007 (UTC)

I'm not confusing issues. NPOV isn't about "who is right," it's about "what views people hold." As I've shown, it's not universally held that GNU is an OS and therefore, stating that it is one, is a violation of NPOV policy. What you're referring to is "GNU Hurd" not just "GNU" anyway, making it "an operating system called "GNU Hurd." I don't know anyone who refers to a Hurd system as just "GNU," nor "GNU OS." -Nathan J. Yoder 21:19, 1 April 2007 (UTC)

That's probably because nobody actually runs Hurd. Regardless, the article text currently states that GNU is an ongoing and unfinished project, so it's no more sensible to nitpick whether the end result is defined correctly than it is to argue that SETI isn't about contacting aliens because no-one's answered yet. If you want to introduce a well-sourced section to the article disputing whether or not GNU is really an operating system or not then knock yourself out; existential arguments on free software talk pages aren't really my bag. Chris Cunningham 21:31, 1 April 2007 (UTC)

You're missing the point. Regardless of how sensible you think it is, it's still disputed and thus isn't NPOV. NPOV issues aren't resolved by championing the "right side" in the article and justifying on the basis of an added section for disputes. There would be a big revert war if I were to just modify the intro paragraphs so that it doesn't state outright that it's an OS, which is why I want to come up with an NPOV wording BEFORE making changes. That analogy doesn't work; SETI exists exclusively for that one purpose and couldn't possibly be confused as having another purpose, plus they're not making a product so there's no name confusion. -Nathan J. Yoder 01:29, 2 April 2007 (UTC)

How exactly isn't GNU an OS? Or you maybe make a confusion between GNU toolchain that is used by other OSes and the entire GNU that's an OS in construction, if you are curious you can run a GNU Live CD with Hurd kernel, yeah it's crappy (heck it might not even boot on your machine and it definitely wouldn't run your games, but that doesn't make it less of an OS) -- AdrianTM 02:02, 2 April 2007 (UTC)

I've already specified how it isn't and I'm not really going to repeat it at this point. Instead of arguing it anymore, I will simply stick to the point that the article is not WP:NPOV compliant. The question then becomes how to phrase it properly and what citations become necessary. If we are to provide citations for both views (aside from the official FSF/GNU pages), how should this be done? It's not a big enough issue that, AFAIK, there would be a newspaper article written on it. I'm thinking of perhaps a single reference with examples of threads from newsgroups and blogs--simple induction. -Nathan J. Yoder 20:09, 2 April 2007 (UTC)

So you want to push opinions from newsgroups and blogs into an Encyclopedia? -- AdrianTM 21:51, 2 April 2007 (UTC)

I'm not trying to "push" anything. As it stands now, the article is "pushing" a one sided view, I'm trying to correct that, to comply with NPOV standards. Newsgroups, blogs and other sources demonstrate the existence of this other view, do you have better sources for this? Wikipedia can and has used both of the sources in other articles and for good reasons too, so I don't see what your opposition here is, except to keep the article only showing your personal view. Even this very article uses a mailing list, newsgroup and blog as references. Please, remove those from the article right now if you don't think they're good sources. -Nathan J. Yoder 00:31, 3 April 2007 (UTC)

"Newsgroups and blogs" are not reliable sources, please read what WP:NPOV says "All Wikipedia articles and other encyclopedic content must be written from a neutral point of view (NPOV), representing fairly and without bias all significant views that have been published by reliable sources." (my emphasis) Bring other sources and we'll talk about that. -- AdrianTM 00:46, 3 April 2007 (UTC)

And I agree with you if this article uses newsgroups and blog as references there should be removed, but 1. don't ask me to do this because I don't owe you anything 2. some references might be valid in context, if RMS or Linus post on blogs or newsgroups those are their opinions and might be quoted as such, however if jerk3446777 says in his blog that he doesn't consider GNU an OS then that can be safely ignored by an Encyclopedia, as it should... again bring the quote and we'll discuss on it, I can't discuss theoretically about a quote I didn't see. -- AdrianTM 00:51, 3 April 2007 (UTC)

I'm not quoting anyone, I'm simply demonstrating that the contrary view is common. I've already provided a link and search term to find more, please read back in the conversation before posting here asking me to repeat myself (you already did this once before). This isn't just one person, as I've already demonstrated. I'm not sure why you consider sources clearly demonstrating various different people disagreeing with the view unreliable, are you suggesting they were forged by a small group of people intending to look like many? Let's consider the opposite here: where are your sources demonstrating the commonality of the view that it IS an OS? You have the official website, but that's about it.

If you don't provide a quote what are we discussing about? -- AdrianTM 02:15, 3 April 2007 (UTC)

I should also note that you seem to have contradicted yourself. You think that they aren't reliable sources, except...when they aren't. There's nothing inherently unreliable about blog, mailing list and newsgroup postings--or even search engine searches (which are usually used to demonstrate notability of something), and as such, having not even read the links I provided, I find your objection very odd. -Nathan J. Yoder 02:09, 3 April 2007 (UTC)

Please provide the links, do you want me to beg on the knees? I don't think I contradict myself, I said that in general they don't have a place in an Encyclopedia except of course if the persons that published the blogs are notable for the article and I gave example RMS and Linus, versus anonymous_jerk. If the quotes you provide (if you do provide) are from somebody like Linus and they are current and relevant then they should be used. -- AdrianTM 02:15, 3 April 2007 (UTC)

Have you seriously not encountered many people with this view? I did provide a link to a thread above, along with my search, as I said. How about an MIT professor of applied mathematics or one of the Slash (Slashdot code) and MacPerl programmers? (this guy and this guy). I found threads from mailing lists and newsgroups with them discussing it (one even mentioning that it's a common view). I can find them again if you think they're 'qualified.' -Nathan J. Yoder 02:35, 3 April 2007 (UTC)

It's irrelevant what opinions I've encountered, please bring the quote forth and we'll decide if we can use it otherwise we blow only hot air over here. I'll tell you from now what criteria _I_ will use to decide if the quote is good or not:

1. . is the opinion pertinent and relevant to the subject
2. . is the opinion current
3. . who expressed the opinion and in what context.
   

Maybe I forgot something, but that's the basic stuff when deciding to put something in Wikipedia. So think about these 3 issues and see if it's worth to trouble other people with the issue, if you think it is don't hesitate to bring up the quote... don't ask for premission from anybody. -- AdrianTM 03:12, 3 April 2007 (UTC)

Obviously this is relevant to the article...it's the article for GNU. They are current, but that's irrelevant anyway, because Wikipedia is not in the business of excluding historical information. I posted the people in question because the main question here is who expressed it, not the other things. I can look them up *again*, but I'd rather see what you and others think of the people first before I expend more effort. So, do you consider such people to be 'quotable' on this subject? -Nathan J. Yoder 05:01, 3 April 2007 (UTC)

I will cease to discuss this until you provide the quote. -- AdrianTM 12:25, 3 April 2007 (UTC)

The example give was here. I've addressed why it's invalid given the context already. There are few notable sources who agree with the user's declaration and few who dispute that "GNU" can be defined as "an operating system composed entirely of free software", so at any rate it's a minority position and disputing NPOV based on them is psychological projection. I wouldn't advise losing too much sleep over this. Chris Cunningham 13:00, 3 April 2007 (UTC)

No, you haven't addressed why it's "invalid given the context." You stated your opinion on the matter, you didn't address why the quote isn't in this context, nor why it shouldn't be included beyond you simply disagreeing with it. I can't assume good faith with you Adrian, when you keep refusing to address the things I say and ignore the link I've already pasted. I'm not going to quote an entire conversation in here, so please click the inks instead of complaining. Thumperward's position that this is a 'minority position' is unsupported. I've asked for sources beyond the GNU/FSF websites themselves demonstrating the extent of people who DO consider it a "GNU OS" (which people rarely call it that--the only people I've seen do it are avid FSF/GNU supporters). You could, however, make more false analogies.

Here are three more threads (you need to scroll down to get to the GNU OS stuff): (1, 2, 3)

-Nathan J. Yoder 06:32, 4 April 2007 (UTC)

I think the point's been adequately made. If the opinions of random people on Linux kernel discussions made eight years ago (when no GPL kernel other than Linux could actually boot and run programs) are the best source you can come up with, there's little point in continuing this conversation. I'm personally disinterested in spending 50% of my time on Wikipedia trying to tell the GNU and anti-GNU camps that they can't define each other's projects. Chris Cunningham 08:04, 4 April 2007 (UTC)

An MIT professor and Slash+MacPerl programmer aren't random people. It's as if you're deliberately ignoring what I've already said. I don't know why this being eight years ago matters. As I stated, at worst it's historical and that doesn't preclude its inclusion. One was also just 6 years ago, but ignoring that makes it go away. Your idea that their views have changed is pure speculation and your own opinion, which is not a valid reason for exclusion. You seem intent on making excuses to exclude it. You also seem to show no interest at all in verifying that this is, in fact, a minority view, despite you knowing that practically no one says "GNU OS" (what does that refer to anyway, Hurd? MULTIPLE OSes? [which would make 'GNU OS'--singular--invalid]). Nathan J. Yoder 08:54, 4 April 2007 (UTC)

Just 6 years ago? You must be joking or you have no idea how computing field works, it's like bringing a 6 years old quote that Vista is not an OS. -- AdrianTM 13:37, 4 April 2007 (UTC)

I guess I was right, you're not honestly interested in the quotes, now that I've provided them. Did you even bother to read them? So are you suggesting that it was valid to say 6 years ago? If not, you have no valid point. You're also contradicting yourself. First you said that if it's an OS "in construction" that it's still an OS. Now, apparently, if it's in construction, it's NOT an OS. How is this analogous to Vista when GNU Hurd was in the planning for decades and was in development since 1990? Your analogy makes no sense.

I seriously hope you're proud of wasting my time. My intuition was right--and I was right in the beginning to ask you to address the people I listed before I got the quotes again. *sigh* I did just so you couldn't say "but he didn't do it." -Nathan J. Yoder 00:34, 5 April 2007 (UTC)

I think I explained pretty well why we need current info, not old info that doesn't apply anymore. I said that before that current info is important, you can accuse people around of whatever you want, but if you come up with 6 year old quote to prove that something is or isn't now you'll not going to very successful, besides I look at those quote most of the people talk about GNU tools not about whole GNU. -- AdrianTM 03:19, 5 April 2007 (UTC)

It is current. And no, you never explained that. You're just looking for an excuse and the fact that you didn't address anything I just said, including your own contradiction, makes me wonder whether you'd accept any quote for inclusion regardless of who said it and when. So what's your reason for believing that everyone suddenly changed their minds, again? Also, what's your reason for excluding historical information, again? You never justified it, you just compared an OS that was in development for 17 years to an OS in development for only a few years. Besides, the fact that other people weren't talking about doesn't negate the fact that others DID talk about it. -Nathan J. Yoder 04:16, 5 April 2007 (UTC)

"It is current.".... "just 6 years ago" Enough said. -- AdrianTM 04:50, 5 April 2007 (UTC)

The article begins by saying "GNU" is an OS, and provides a link to the article "Operating Systems" which has a concise definition of the concept; unfortunately the description given there does not fit GNU in any way.

It is true that GNU began as an attempt to clone UNIX and that the project eventually released the Hurd, the GNU kernel; but the GNU applications work with any POSIX compliant kernel - why else is Stallman always insisting that Linux be called "GNU/Linux"?

That first paragraph should be rewritten.

Can't you read a little bit above? Why open a new discussion? -- AdrianTM 17:47, 1 June 2007 (UTC)

Regardless of anyone's opinion here, the only definition of "operating system" that matters in this context is POSIX. GNU provides the kernel, C library, core utilities and shell required to call it's self a POSIX operating system. It's about as simple as that. Noahslater 12:55, 15 November 2007 (UTC)

GNU is either not an OS or it is incomplete

An OS requires a working kernel. The GNU project does not have a working kernel. This means either that GNU is incomplete or that it is not an operating system. You can't have it both ways. Either we describe it as incomplete, or as an unfinished project, or as a collection of userland packages. It is POV to claim it to be an operating system when it does not have its own kernel. If people insist on calling it an OS, it should be redirected to GNU/Linux, because that is the primary way people solve the problem of its incompleteness. The reference given says that GNU is incomplete because the kernel is incomplete. Yworo (talk) 13:42, 30 June 2010 (UTC)

See also Talk:GNU#Incomplete above. Gwen Gale (talk) 14:10, 30 June 2010 (UTC)

I don't mind at all that my one-off edit was removed, it was only meant as help. I'm neutral as to any outcome, my comments here are meant to help editors think about sourcing and other policies. However, if editors keep going back and forth over this, rather than reaching a consensus somewhere on this talk page first, I will protect the article from editing. Gwen Gale (talk) 15:53, 1 July 2010 (UTC)

Primary sources

As I am sure most of you are aware, we cannot rely on primary sources (i.e. the GNU project and/or the FSF) for the information in this article. This article appears to be written from a fan's point of view based on claims of the organization producing the product. Yworo (talk) 13:57, 30 June 2010 (UTC)

Some primary sources are allowed if care is taken. There is nothing untowards, say, about quoting gnu.org's website with an attribution to the website in the text itself, so readers know straight off where it came from. Gwen Gale (talk) 14:08, 30 June 2010 (UTC)

There is when the GNU project's claim that GNU is an operating system is taken over the definition of an operating system, which requires a working kernel. Also, it would be better if third-party sources were used, wouldn't it? Yworo (talk) 14:11, 30 June 2010 (UTC)

There is no harm whatsoever with putting their (verifiable as having been made) claim into the article text. Likewise, it would be helpful to cite secondary sources (what you call "third party") which both dispute that claim and support it. Gwen Gale (talk) 15:05, 30 June 2010 (UTC)

I have no problem with that and intend to wait 24 hours before making another change myself in any case. Though I wish all the involved users would join the discussion on this talk page to work out consensus wording. I don't think GNU variants such as GNU/Linux should be used in the lead to justify a claim that GNU is finished, though. Linux was never made with the intention of completing GNU, is not part of the GNU project, and never was. Yworo (talk) 16:05, 1 July 2010 (UTC)

The thing to do here is follow what the sources say. If they don't say the same things, deal with that in the text, too. As for the software, as it happens, I've been using GNU software here and there (on FreeBSD) for years and truth be told, when I was asked to look at this, I wasn't even aware they were still hoping to release a stable OS. Gwen Gale (talk) 16:18, 1 July 2010 (UTC)

Oops, this wasn't where I'd intended to post, but discussing sources does seem the next thing to do. The non-GNU sources I've cited use the word "incomplete". Specifically, they say that GNU as an OS is incomplete because it's kernel is incomplete. I'm not sure that FSF has ever described GNU as a complete operating system, except in the context of "GNU variants" or as a future goal. I'd like to see one reliable source that described GNU as a completed, production-ready, operating system rather than as a project with that goal. Yworo (talk) 16:31, 1 July 2010 (UTC)

It seems to me, the big worry other editors have is only with the wording of the lead. I would think that in the following text, sundry sources and adjectives wouldn't be taken as misleading and would give readers a thorough overview of cited outlooks on the kernel. Gwen Gale (talk) 22:55, 1 July 2010 (UTC)

Dubious

Someone has changed the subject of the article from GNU to GNU/Linux, which has its own article. This change should be reverted. Yworo (talk) 16:56, 30 June 2010 (UTC)

RFC: Is the word "incomplete" adequately sourced for the lead paragraph?

Is the word "incomplete" adequately sourced for the lead paragraph? Yworo (talk) 16:14, 5 July 2010 (UTC)

Background: The incompleteness of GNU has been discussed for a number of years, with (primarily) GNU proponents always removing the word without providing supporting sources for GNU's completeness. Some of the more recent discussions can be found in these section of this talk page:

#Incomplete

#Operating System?

#GNU is either not an OS or it is incomplete

#Primary sources

The proposed version of the lead paragraph is:

GNU (pronounced /ˈɡnuː/) is a computer operating system composed entirely of free software. It is currently incomplete as GNU Hurd, its official kernel, is not yet production-ready, but production use is possible using a Linux kernel or other free kernel. Its name is a recursive acronym for “GNU's Not Unix!” This name was chosen because GNU's design is Unix-like, but differs from Unix by being free software and containing no Unix code. Development of GNU was initiated by Richard Stallman and was the original focus of the Free Software Foundation (FSF).

Is the word "incomplete" adequately sourced for the lead paragraph? Yworo (talk) 16:14, 5 July 2010 (UTC)

From RFC: I agree that if the article is about the operating system, that the term incomplete is appropriate. I'm not close to convinced by the argument that it is incomplete as MS-DOS is. As an aside, I think GNU muddies the definition of an operating system by saying that it includes "applications and developer tools" [2] -- GNU is, for the most part, a collection of userspace tools as another editor mentioned. If the article was about the set of userspace tools that GNU decided to call part of an OS, then I wouldn't support calling it incomplete. However, since it is about the incomplete OS, I, of course, support including incomplete. jheiv ^{talk contribs} 07:11, 8 July 2010 (UTC)

From RFC: While the gnu.org site seemingly uses the term GNU to describe an operating system, the site itself is inconsistent. The "Download GNU" link does not provide an immediately obvious way to download the "GNU Operating System", but instead discusses Linux (or "GNU/Linux", for those who prefer) distributions, and "GNU software" (not including anything resembling an "operating system", just individual components). In my view, "GNU" is a collection, and not an exhaustive or complete one, of mostly user-space code which is part of the "GNU Operating System", or "GNU/HURD".

One has to spend some considerable time looking for a way to install "GNU/HURD". In fact, there is no direct link from the "GNU Software" page (at http://www.gnu.org/software/) to the "GNU HURD" page (at http://www.gnu.org/software/hurd/hurd.html), which seems surprising if "GNU" (the organization) is indeed actively promoting "GNU" (the operating system). I'd note that, just as with the term "free", proponents of "free software" have also overloaded the acronym "GNU" to the point where it's unclear what's being talked about: a project? A set of components? An operating system? I'd suggest that if the intent is to discuss "GNU, the operating system", then the appropriate terminology would be either "GNU/HURD" (my preference) or "The GNU Operating System".

I would describe this operating system, "GNU/HURD" as "purely experimental and not yet suited for production applications", unlike "Linux" (or "GNU/Linux"), thanks to the very unstable state of the HURD kernel. "Incomplete" or "unfinished" seem destined to invite hair-splitting from partisans who will, for example, claim that the Linux kernel is "unfinished" since there is still development work being done on it—a definition of "finished" which a piece of software can only seemingly satisfy when it's past its "end of life". 68.126.206.84 (talk) 23:12, 8 July 2010 (UTC)

Additional references re completeness / incompleteness

Here is a place to list, under the two subheadings provided (added Yworo (talk) 17:10, 5 July 2010 (UTC)), additional references both for and against completeness. Yworo (talk) 16:31, 5 July 2010 (UTC)

Unproductive argument about talk page etiquette

:Please do not move editors entries or otherwise refactor this discussion. Such tactics are not fair, and are prohibited by talk page guidelines. This means you, Cyclopia. Yworo (talk) 16:45, 5 July 2010 (UTC) Your statement is irrelevant wikilawyering. You said explicitly above: "Here is a place to list additional references". This means that you leave other users to use this place to list the references, and this also obviously includes recategorizing them. It's not a talk page comment, it's not like I am refactoring your comments to make you say something you haven't said. So, my move stands. Either this is a place to list additional references, or this is a place where you and only you give your own personal opinion on references. If you edit to explictly state it is the second, I won't move again. If it is the first, it means you leave other users the ability to help doing this categorization of references. --Cyclopiatalk 16:50, 5 July 2010 (UTC) You may not relegate, on your own opinions, another editor's contribution as "irrelevant". You may comment in the place where I put it, and you are of course welcome to call your own references irrelevant. Yworo (talk) 16:55, 5 July 2010 (UTC) It's not your contribution, it's the reference which is relevant but not suitable (not "irrelevant", I fixed that mistake long ago). Now, please answer questions: is this a place where we collectively build a list of reference, or it is your sandbox to collect your own opinion on references? It can't be both. --Cyclopiatalk 16:58, 5 July 2010 (UTC) Then comment after the entry. Refactoring the discussion in this way is counter-productive, it is against talk page guidelines, it is uncivil, and you are edit warring over it. Do it again and I will report you to all the appropriate noticeboard and you will likely be blocked. Yworo (talk) 17:00, 5 July 2010 (UTC) I ask again: is this a place where we collectively build a list of reference, or is it your sandbox to collect your own opinion on references? It can't be both. Please give an answer, because if it's the first, it means it can be edited and that they are not comments. If it's the second, please clarify it in the initial statement, which gives the impression that it is a place where everyone is welcome to edit. There is nothing "uncivil" in refactoring a list of references. What is uncivil are your nonsense templates on my talk page and your threats. --Cyclopiatalk 17:04, 5 July 2010 (UTC) The heading an editor gives to an entry is part of their comment. There should be only the two sections I added. Put an entry under one or the other and the we can discuss relevance in place. Yworo (talk) 17:07, 5 July 2010 (UTC) You are refusing to answer my question above, which is pretty revealing. Yworo, you can't create a public space to add references and then request that we follow your rules like "there should be only the two sections I added". We're not here to play your games, we're here to collectively work to make this article better. So, if you want people to help, and you provide an open, public canvas for people to list references and help selecting them, you have to accept that your own categorization of such references could be contested and that other sections can be added. If instead you want to state your own opinion on references, fine, but do not show it as a public, open canvas. You can't say "We'll work together to categorize references, but we will do it with MY rules!". Either we do the rules together and work together, or you do your own rules, and work alone. For this reason, I am removing my entries from below, since I don't want to legitimate this situation unless you decide a clear position. --Cyclopiatalk 17:14, 5 July 2010 (UTC) I'm not refusing to answer anything. You changed the terms of the discussion I started. You don't get to do that. There are two categories of reference here. Adding two more categories called "Irrelevent" and "Unuseable" or whatever they were is not productive. Moving other editors comments is against guidelines. You are edit warring over talk page refactoring, which isn't permitted. I've been quite clear about it, and given you a series of official warnings with links to the policy on your talk page. Which you didn't read, apparently. Yworo (talk) 17:23, 5 July 2010 (UTC) I know the policy, thank you (I had problems in the past with people removing my comments and therefore I know WP:TALKO). Let us review the situation: You create a paragraph starting Here is a place to list additional references both for and against completeness. - This implies that you want people to add and list more stuff in there. I add another category under which to list references and move one of the reference there, with explanation. You aggressively revert saying that it is refactoring your talk page comments I ask you "well, if so, do you really want people to edit this paragraph and help in categorizing references or is this just a place where you dump how you personally categorize them without further help?" You refuse to answer to this question and you say that we should follow your rules, for some unknown reason, to use the paragraph (and then explicitate that on top) When I refuse to legitimate that and therefore move my edits, you send me another TALKO violation warning, like you were actually owning the paragraph. So? Again, either this is a public place where to help categoriaze reference, or it is your own playground. But it can't be both. Choose one, and everything becomes fine again. No need to extend this lame discussion further, just take your pick. --Cyclopiatalk 17:31, 5 July 2010 (UTC) I think I've made my position very clear. You are being obtuse. That's my final comment on the matter. I've reported you for vandalism, and at Wikiquette alerts. Go have a discussion on the matter with uninvolved editors. This is my last response to your studied obtuseness. Yworo (talk) 17:36, 5 July 2010 (UTC)

Supporting completeness (or being finished, usable, etc.)

• [3]: "The current state is still far from being on par with Debian's established Linux ports, but it is mostly up to date and reasonably usable." added by Yworo (talk) 17:19, 5 July 2010 (UTC)

That's a status update from the project itself. We need third-party sources. Yworo (talk) 16:51, 5 July 2010 (UTC)

Supporting incompleteness (or unfinished, unusable, unstable, etc.)

• Hillesley, Richard. "GNU HURD: Altered visions and lost promise", June 30, 2010. See especially page 3: "Nearly twenty years later the HURD has still to reach maturity, and has never achieved production quality." ... "Some of us are still wishing and hoping for the real deal, a GNU operating system with a GNU kernel." (added by Yworo (talk))

(restored from history) "(Does not say it is incomplete, it says that it "has still to reach maturity" and "never achieved production quality"). --Cyclopia^talk 16:36, 5 July 2010 (UTC) - Moved to "Relevant but not suitable" for reasons above. --Cyclopia^talk 16:50, 5 July 2010 (UTC)"

• (from gnu.org itself) hurd: "The GNU Hurd is under active development. Because of that, there is no stable version." and "Debian GNU/Hurd is currently under development and available in the unstable branch of the Debian archive." added by Yworo (talk) 21:02, 5 July 2010 (UTC)

• (from sourceforge) GNU Hurd at Sourceforge Page: "Note that this is an "unstable" distribution, in that it is not currently read[y] for use in a production environment." Note that earlier in the paragraph is says that the packaging is quite stable.added by Yworo (talk) 21:07, 5 July 2010 (UTC)

Yworo, if you want to cite sources to support putting the word incomplete in the lead, you'll need to cite only sources which use that word.

"Everyone," please stop refactoring talk page posts, even in good faith. Gwen Gale (talk) 20:09, 5 July 2010 (UTC)

Since Cyclopia seems to want something on the order of "It has been described as...", I am now in favor of listing the different variations on the theme of incomplete, unfinished, unstable, etc. Yworo (talk) 21:02, 5 July 2010 (UTC)

protected

Owing to ongoing edit warring, I have protected the article from editing. Please use this talk page to reach consensus. Gwen Gale (talk) 20:26, 5 July 2010 (UTC)

Gwen, I was simply implemented what Cyclopia has said was acceptable to him in the discussion here, earlier today. He said he would have no problem using wording like "It has been considered...". I used "It has been described..." Anyway, at least you protected the "right version" :-) Yworo (talk) 20:29, 5 July 2010 (UTC)

Truth be told, I most likely protected the non-consensus version, the timing was happenstance (meanwhile, one other editor is not consensus). Gwen Gale (talk) 20:39, 5 July 2010 (UTC)

Lawrencekhoo seems to be onboard with the change, and IMO much improved what I was trying to include. Really only two editors have been opposed, Grandscribe and Cyclopia, and the latter's objection was to not including wording to the effect that "It has been considered..." which I added. Grandscribe is well-known for going on binges of changing Linux to GNU/Linux in multiple articles whether it is appropriate or not, and his opinion in this case can be dismissed as entirely part of his POV. Yworo (talk) 20:44, 5 July 2010 (UTC)

Let's hear from them then. Gwen Gale (talk) 20:49, 5 July 2010 (UTC)

The current wording is quite fine with me, because it says that it has been described and not that is, a very important distinction. Apart from that, I would probably move most of it from the lead and create a "Current status" paragraph where we can elaborate on the issue at length. --Cyclopia^talk 21:58, 5 July 2010 (UTC)

The problem with that is that the uncontested implication in the lead paragraph sentence is that GNU is a completed operating system, which is not the case. The fact that this is contested needs to be in the lead paragraph, per WP:LEAD which states that the lead has to adequately summarize the content of the article, and WP:NPOV which states that all significant viewpoints must be given balance. We have yet to see any citation which call GNU a finished, completed, production-ready OS. To imply that it is by simply calling it an "operating system" without qualification is to give an inaccurate impression in the lead paragraph. If that statement is going to be the first sentence, informed conflicting opinions must be in the second. Yworo (talk) 01:19, 6 July 2010 (UTC)

• GNU is a completed operating system, which is not the case.: Find an objective definition of "complete" for OS. You've been asked this several times, you never provided an answer. The only source that talks of "incompleteness" is a mere opinion piece. Complete/incomplete are misleading statements, unless they have a clear definition that applies.
• We have yet to see any citation which call GNU a finished, completed, production-ready OS. : Production-ready is *different* from complete, and not production ready is different from incomplete. A system can be "complete" (whatever it means) feature-wise but be very unstable, for example. There are two related issues in this respect:

• Is a GNU/Linux or GNU/kFreeBSD considered a GNU system the same, or not? If yes, there are GNU(/Linux) systems considered production ready. This ambiguity should be explained clearly.
• The key problem is the status of the official kernel, GNU Hurd. So, issue above notwithstanding, the real issue is the status of GNU Hurd. I suggest that we should cover the issue shortly here and do it in full in the GNU Hurd article.

• I want to clarify that nobody denies here that GNU (with Hurd as a kernel) is an unstable and absolutely non-production ready OS. What I am worried is the word "incomplete", which is misleading because it has the implication that GNU "lacks pieces", which is false: the pieces are all there, they are simply of bad quality and stability. --Cyclopia^talk 12:54, 6 July 2010 (UTC)

I understand all the points and arguments you bring up, but:

• When you ask for definitions as you are, you are asking me to do original research. I don't have to because I have sources that use the words, "incomplete", "not mature", etc.
• Yes, production-ready is different. But you still don't have a reference that calls GNU complete. If GNU is complete, you should have no problem finding one, along with the initial release date of the completed, production-ready operating system. If you can find a source describing the status, sucj as all major subcomponents present or some such, that can be contrasted with the other opinions.
• No, GNU/Linux or GNU/kFreeBSD is not the GNU operating system. FSF defines GNU as a project with the goal of creating a complete operating system. Components which are not part of the project cannot be called "GNU". It's intellectually dishonest. If you'd like to merge this article with GNU variants and call it GNU systems we could discuss that.
• Yes, I know the problem is the kernel. But for the GNU project to be complete as described and self-sufficient, the kernel must evolve to be production-ready.
• The word "incomplete" is in one of the sources, as is the other verbiage. When there are major disagreements about the status of something, especially when they are so major as to bring into question the definitions used in the first sentence, don't you think these disagreement need to be clearly stated up front?

Ok, let me answer (I am happy we're back talking in a civil manner by the way!)

• Again, you define "incomplete" as a synonim of "not mature" or "not production ready". It is not. The Merriam-Webster main definition of "complete" is having all necessary parts, elements, or step - Therefore if GNU is "incomplete" it means that it lacks a "necessary part, element or step". Now, if we say that it is incomplete, the first thing that comes up in the mind of the reader is: which necessary part,element or step is lacking? That is why it is misleading. A wrecked and unusable car is not necessarily an incomplete car. If I say "My car is not well functional" and I say "My car is incomplete" I say two different things.
• I won't bring you also a source saying that GNU is "complete", because it is also another nonsensical statement. I would have objected to "complete" as well as to "incomplete": even if "complete" can have a reasonable minimal definition that could be met by GNU (it compiles, boots and runs programs, the essential aims of an OS), it would be misleading to define it so.
• Ok. I am neutral on the issue, I was just asking because it is something quite unclear to me (even the FSF maintains a degree of ambiguity on that).
• Agree.
• The word is in the source, but the source is just an opinion column, and I don't think it is suitable for giving definitions (while it is perfectly OK if we talk of the public reception/description of the thing). About the disagreement, I am unsure of what disagreement you talk about. --Cyclopia^talk 17:00, 6 July 2010 (UTC)

This issue has been brought up repeatedly by other editors and no changes have ever stuck. Sometimes agreements have been reached. At one time "complete but unfinished" was in the lead sentence. But these compromises have always been removed later by editors I can only assume are GNU proponents. Let's make sure the reader clearly knows the status of the product being produced by the GNU project up front. It took more that two decades just to get to version 0.2 of Hurd. 0.3 has been supposedly in the works for quite some time and is still a no show. Why would we imply a released OS in the first paragraph without making it clear that that is an illusion. Yworo (talk) 13:23, 6 July 2010 (UTC)

Clearly there are other ways to resolve the problem. We could have an opening sentence that reads "GNU is a computer operating system which has been under development since 1983 but which has not yet had its first stable release." Would you prefer something along those lines? Yworo (talk) 13:23, 6 July 2010 (UTC)

I would absolutely agree with that, because it is factual and verifiable. Let me know what do you think about building a short paragraph that details the status of the OS. --Cyclopia^talk 17:00, 6 July 2010 (UTC)

Well, if that can be the lead sentence, then I have no problems with the opinions being moved later in the article. No need to keep them in the lead if the "not yet had its first stable release" can be in the lead sentence. I'd like to keep the info about it being usable in production with the Linux kernel or another kernel there. It is only fair to include that in the first paragraph. It then seems to me there should be a paragraph explicating that that userland parts are indeed finished and robust and distinguishing that from the state of the Hurd kernel. I would like to see a date at which the major userland components were all production-ready, rather than the fairy-tale-like "long ago". Yworo (talk) 17:09, 6 July 2010 (UTC)

Agree completely. --Cyclopia^talk 17:13, 6 July 2010 (UTC) Retracted - apparently I didn't understand correctly the terms of the agreement, I didn't think that "the lead sentence" or "the first two-three sentences" would have made a large difference. I agree in having "not yet had its first stable release" in the first 5 sentences of the article, and even in the first if there is an elegant way of doing that, but I disagree in having it as a milestone. --Cyclopia^talk 22:30, 6 July 2010 (UTC)

One vs two sentences

Can you explain this diff? My problems with it are:

• The edit didn't change anything in the meaning, it simply made the sentence more readible by breaking it in two. If you disagree, could you explain why?
• You can't ask for "no other changes" and "I didn't agree" is not a reason to revert in itself -the article is not yours. Of course you are entitled to revert, but please explain why you disagree with the edit. --Cyclopia^talk 21:01, 6 July 2010 (UTC)

You "agreed completely" to that single sentence. However, further input from you and other editors is very welcome. Gwen Gale (talk) 21:05, 6 July 2010 (UTC)

Did I? If so, I totally missed the part that fixed the thing to be absolutely in one and only one sentence. That said, the sentence as it is now is excruciatingly long. Anyway, we can go further, since the lead as it is now is a bit messy. I propose we can take example from other -apparently quite more mature- OS articles like Linux or FreeBSD:

FreeBSD is a free Unix-like operating system descended from AT&T UNIX via the Berkeley Software Distribution (BSD). It has been characterized as "the unknown giant among free operating systems". It is not a clone of UNIX, but works like UNIX, with UNIX-compliant internals and system APIs. FreeBSD is generally regarded as reliable and robust.

Linux [...] refers to the family of Unix-like computer operating systems that use the Linux kernel. Their development is one of the most prominent examples of free and open source software collaboration; typically all the underlying source code can be used, freely modified, and redistributed, both commercially and non-commercially, by anyone under licenses such as the GNU General Public License.

OpenBSD is a Unix-like computer operating system descended from Berkeley Software Distribution (BSD), a Unix derivative developed at the University of California, Berkeley. It was forked from NetBSD by project leader Theo de Raadt in late 1995. The project is widely known for the developers' insistence on open source code and quality documentation, uncompromising position on software licensing, and focus on security and code correctness. The project is coordinated from de Raadt's home in Calgary, Alberta, Canada. Its logo and mascot is a pufferfish named Puffy.

The structure of these two is a bit different but it seems setted on a reasonable pattern of:

• Sentence 1: what is the thing
• Sentence 2-4: Main features/history of the OS

So what about something like :

GNU is a free Unix-like operating system developed by the GNU project since 1983. In its official version, with the GNU Hurd kernel has not issued a stable release yet. GNU/Linux, which substitutes the Linux kernel to the GNU Hurd, is however generally considered production ready; also other free kernel replacements exist with varying degrees of stability. GNU is composed wholly of free software: all the underlying source code can be used, freely modified, and redistributed, both commercially and non-commercially, by anyone under licenses such as the GNU General Public License. To paraphrase Richard Stallman, leader of the GNU project, the GNU system is a technical means to a social end.

which in my opinion prioritizes better:

• Sentence 1: what is GNU, who develops it, since when
• Sentence 2: It is not stable yet
• Sentence 3: The GNU/Linux thing
• Sentence 4: One of the defining characteristics of GNU
• Sentence 5: Why does it have such characteristics, in the words of his creator

Any comment? --Cyclopia^talk 21:27, 6 July 2010 (UTC)

I specifically required that the words "operating system" be qualified in the same sentence with the lack of release status. You completely agreed to it. I will not agree to any version that does not have the qualification in the lead sentence as I presented it. I explained my reason thoroughly. I've now having a hard time assuming good faith about your motivations. Yworo (talk) 21:39, 6 July 2010 (UTC)

If you did explain the reason thoroughly for it to be in the very same and one sentence, I sincerely missed it, and I would like to see it. I agreed to have it in the lead, I didn't think that having one or two sentences was an essential part of the thing. That said, I am baffled about your AGF comment -could you elaborate what my "bad faith" is, because I really can't see what kind of agenda do you think I pursue by splitting a sentence into two. We agree on the facts, we just don't agree on how we present them. It would be appreciated if you can stop reacting with such aggressivity to whatever I say/do and argue in a pacific, civil way. --Cyclopia^talk 21:57, 6 July 2010 (UTC)

The other operating systems are different. There is no question that they have had a stable release. The whole point I have been arguing is that the term "operating system" is misleading in GNU's case without qualification. Yworo (talk) 21:42, 6 July 2010 (UTC)

If this is your point, your point is wrong. To say that GNU is an operating system is the plain truth and it is not misleading in any way. To say that GNU is an operating system which also never saw a stable release is also the plain truth and it is simply more precise. To say that GNU is an operating system which is fully free software and also never saw a stable release is also the plain truth and it is even more precise. There is no need to add specification layers in a single sentence. We can use more than one sentence to make it easier to read. --Cyclopia^talk 21:57, 6 July 2010 (UTC)

And two dependent clauses does not make a sentence "excruciatingly long". Sheesh. Yworo (talk) 21:45, 6 July 2010 (UTC)

Well, I disagree. To me it's quite a clumsy sentence. Not a big deal, but I don't see why you insist so stubbornly on that. --Cyclopia^talk 21:57, 6 July 2010 (UTC)

I used the phrase "lead sentence" over and over again in my postings. Either you don't read very closely or are being disingenuous. I'll give you the benefit of the doubt and ask you to read more closely in the future. Yworo (talk) 22:00, 6 July 2010 (UTC)

None of the two. I simply thought (my fault) that "lead sentence" was a general statement to say that it has to be in the very beginning, not that it had to be a totally strict requirement. To be honest I really didn't care about that until I read the sentence in the article and I just thought "oh well, it reads better if we split it", and here we are. So, if you require them to be in the same sentence, I disagree (and, if you want to be reassured, it is for matter of presentation purposes and lead structure, nothing else: I wholly agree on the fact that the unstable status of GNU has to be presented, say, in the first 5 sentences at most -again, your innuendos are quite annoying, could you please explain clearly and once for all what do you perceive I'm doing in bad faith?). Now, why do you want them to be absolutely in the same sentence? --Cyclopia^talk 22:08, 6 July 2010 (UTC)

You've just given me an example. I've already explained and you insist I have to explain again. So you disagree with my explanation. Too bad. It's still my explanation and I disagree with your dismissal of it. You are playing semantic games and I am done. I am holding you to your agreement. If you think that adding the second "which" that got dropped helps the readability, go for it. Yworo (talk) 22:11, 6 July 2010 (UTC)

I've already explained and you insist I have to explain again. - Where? Just link the diff to it. I haven't seen it, honest!

So you disagree with my explanation. Too bad. It's still my explanation and I disagree with your dismissal of it. - No, I disagree with an action of which I have not seen an explanation. That said, if we disagree, the article is not yours and you can't expect to simply put it at rest this way.

You are playing semantic games and I am done. - No idea of what games you're talking about. Please stop being aggressive.

I am holding you to your agreement. - First, you can't hold me to anything. Second, I never agreed to have this in a whole clumsy sentence. I agreed to have this in the lead. Since it seems that this is for you a fundamental issue, while I thought it was a very secondary thing, I retract my agreement above and clarify it. Gwen Gale, please reprotect the page -issue is apparently not solved. --Cyclopia^talk 22:25, 6 July 2010 (UTC)

My point proven. Yworo (talk) 22:27, 6 July 2010 (UTC)

Your point what? Yworo, you're bordering on incivility and article ownership. Please explain. --Cyclopia^talk 22:30, 6 July 2010 (UTC)

I think I've already explained quite thoroughly. Nor have I been uncivil or professed article ownership beyond asking you to stick to the agreement you made. Clearly, not sticking to an agreement is difficult to consider as a shining example of good faith. Especially after the long and detailed explanations I made with the inclusion of "lead sentence" at least five times over the course of the discussion. Yworo (talk) 22:38, 6 July 2010 (UTC)

I already explained that I made a mistake: I didn't think that the "lead sentence" was to be taken so literally, and I was wrong. My fault, as said above and in my retraction. I think it's difficult being more honest than so. Now, the closest thing I've seen to an "explanation" of this requirement is when you say: The whole point I have been arguing is that the term "operating system" is misleading in GNU's case without qualification. above. We can discuss about that -I disagree it is misleading- but even if we agreed, this in no way implies that this qualification has necessarily to appear in the same sentence. That's why I didn't consider it a proper explanation -if you had another one and I missed it, why don't you just link the diffs to me so that we can reason? --Cyclopia^talk 22:45, 6 July 2010 (UTC)

And I don't think the sentence is too long, so I dismiss that as a valid reason for splitting it. Yworo (talk) 22:47, 6 July 2010 (UTC)

Ok, this is perfectly fine. But "not thinking that the sentence is too long" does not mean "thinking that it can't be two sentences". What is the problem with splitting it? --Cyclopia^talk 22:49, 6 July 2010 (UTC)

Because "operating system" implies something that works for the average, general reader. And GNU most likely won't run for such a reader. You might be able to get it up and running on the right hardware on the first attempt; hell, even I might be able to do it. But the average reader is going to trash their computer if they don't understand about the incompleteness (missing driver, instability). Splitting the sentence makes it seem incidental rather than deep-rooted. Yworo (talk) 22:54, 6 July 2010 (UTC)

(I tried a compromise in my last edit, see what you think of it). Good to know where you are standing. Let me tell you: I almost completely agree (heck, have to be careful now saying that!) with most of your post. What I disagree what is the first and the last sentences.

• Because "operating system" implies something that works for the average, general reader. - This is a very peculiar POV which has no roots in the definition of operating system. Our article says, for example: An operating system (OS) is a set of system software programs in a computer that regulate the ways application software programs use the computer hardware and the ways that users control the computer. For hardware functions such as input/output and memory space allocation, operating system programs act as an intermediary between application programs and the computer hardware,[1][2] although application programs are usually executed directly by the hardware.. GNU Hurd does all of that (and it is sourceable). I have never seen operating system used in a meaning that implies good usability for the average person (I'd say that if it was the case, almost no OS truly exists ).
• But the average reader is going to trash their computer if they don't understand about the incompleteness (missing driver, instability) - This is very true (honestly, the average reader probably couldn't even start the install) but it is not our concern: probably the same reader would be in deep trouble trying to install Minix or Plan 9, yet we are not warning people about them. We're not a guide to help people (not) install weird OSs on their computers.
• Splitting the sentence makes it seem incidental rather than deep-rooted. - I agree that the disastrous state of Hurd is deep-rooted enough to warrant a mention in the first sentences. But it is not required to be in the very first sentence. We don't write that FreeBSD is a good and stable and usable OS in the first sentence, we simply declare it is an OS. The status of the OS is important, and in the case of GNU is even more, given the "failed promise" status of it, but to insist to have it in the first sentence is unneeded, see above. --Cyclopia^talk 23:05, 6 July 2010 (UTC)

So, is your complaint not really sentence length? Because removing the "in development since 1983" and using "which" followed by the original agreed wording is an equally short sentence to using a semicolon. Surely that is short enough? Yworo (talk) 23:09, 6 July 2010 (UTC)

My complaint is in general the lead, which is a bit messy. I proposed an alternative (first half of the) lead above. Shortening the first sentence is absolutely important, but it is only a part of it. I would prefer a more rational structure consistent with other articles, which declares first what is GNU, who does it and when it started -and whose subsequent sentences declare the essential features of it: its status, its aim, etc. I am going to try another compromise now, let me know what do you think, it seems we're heading in the right direction despite a bit of reciprocal misunderstandings. I would appreciate if you can answer to the above comments, to know what you think about. --Cyclopia^talk 23:16, 6 July 2010 (UTC)

I've considered it quite thoroughly. Having not had a stable release is of equal importance with it being (or claiming to be) an operating system. Thus it should be mentioned in the lead sentence. Your other prioritizations seem reasonable to me. Yworo (talk) 23:17, 6 July 2010 (UTC)

Well, I think you will like my last edit in this respect. I'd say we can settle on that for now, take a deep breath and later discuss on how to restructure the lead, maybe with the help of some other editor to help to unblock debates like our last ones and reaching proper consensus. --Cyclopia^talk 23:22, 6 July 2010 (UTC)

Sigh. Now what is wrong? --Cyclopia^talk 23:24, 6 July 2010 (UTC)

Well, no I didn't. Why would the language be good when I proposed it but not good now. "has not yet had its first stable release" is much more accurate (and telling), than "is not yet stable". Why do you want to gloss over this? Yworo (talk) 23:25, 6 July 2010 (UTC)

Because this one is better (shorter sentence and fits your criteria) and it didn't occur to us before? It's not like solutions found once are crystallized forever -this is a wiki, after all. How putting that it is not stable as the very first three words of the article can be interpreted as "gloss" makes no sense whatsoever. --Cyclopia^talk 23:33, 6 July 2010 (UTC)

It doesn't fit my criteria. It's a less accurate description and leaves out the element of time, which you already butchered significantly by removing the start date from the sentence. I'm asking Gwen to re-instated the agreed upon sentence. Give you an inch and you try to take a mile. I am not longer willing to remove the project start date and the length of time it has take to not issue a first stable release. Yworo (talk) 23:37, 6 July 2010 (UTC)

Start date re-added. Now? --Cyclopia^talk 23:52, 6 July 2010 (UTC)

What I think is that you are getting dangerously close to breaking 3RR (partial reverts) and that I am not doing any more editing today. I'll take it up with Gwen tomorrow. Yworo (talk) 00:04, 7 July 2010 (UTC)

So now you consider trying to honestly compromise to be edit warring? WP:OWN, WP:OWN... --Cyclopia^talk 00:11, 7 July 2010 (UTC)

Uh, when you keep reverting to language I've clearly disagreed with, simply in a different arrangement, I'd say you are the one with ownership issues. I'm not the one reverting now, am I? Yworo (talk) 00:16, 7 July 2010 (UTC)

I've asked at WP:RPP for full protection of the page (making it clear that they can revert to your version and protect it if they choose so -I am not requesting it to protect my version, just to be sure, I simply want this to be settled here before we go editing each other again). Now: You disagree with my language, and I disagree with yours. Fine. I prefer an initial sentence less focused on the status and more readable than the one of this afternoon. Now, to settle this, you've put on the table some requirements you wanted to be met:

• You wanted the GNU not yet stable status to be included in the lead sentence. This and this satisfied this request.
• Then you disagreed because "it was not the same language" ([4])
• When asked what you were disagreeing with in detail, you talked about the issue of time.
• Here, to acknowledge your requests, I put back the time thing them, putting back also the date in which it was started.

So, if you are honest in your arguments, you can for sure say "Well, ok, it is an improvement but I still think it needs tweaking for this or that..." or something like that, letting me understand what you think can be made better. We need to compromise. Statements like "I agreed on one sentence,not two; no other changes" or "reverting language I've clearly disagreed with" are strong hints that you are attempting to own the article (Wikipedia:Own#Examples_of_ownership_behavior Actions #1, On revert #2). I came here discussing all my attempts of compromise and I've tried always to compromise with your requests. --Cyclopia^talk 00:33, 7 July 2010 (UTC)

References

Request for Gwen

Could you please re-institute the full agreed upon wording of the lead sentence. And could you please discuss with Cyclopia just why the sentence was good in discussion but no good any longer. I disagree that it is too long; and Cyclopia's edits make clear that length isn't really the issue anyway. He seems to have a problem with making the status too explicit. I have no more patience for his argument style and seriously doubt his good faith. Someone else needs to pursue his reasons and real motivations. Yworo (talk) 23:30, 6 July 2010 (UTC)

I [...] seriously doubt his good faith. Someone else needs to pursue his reasons and real motivations. - This is very close to a personal attack and I ask you to retract this. My reasons and motivations are only those of having a reasonably good lead, of avoiding misleading statements in the lead and in general have a NPOV, reasonable article. Yworo is the one pushing the odd POV that an OS can't be qualified as such if it's not ready for the general user. In the meantime I ask Gwen to re-protect the article -all my attempts to reach a compromise have failed and it is better to protect it while seeking consensus from third parties. --Cyclopia^talk 23:37, 6 July 2010 (UTC)

No, it's not a personal attack. Anyone can see you've reneged on your agreement and that you are trying to weaken the language. It's about the edits, the interminable argumentation, reasons changing, etc. Seems like civil POV pushing to me. Killing with politeness while continually watering down language are the first two bullet points, in case you don't bother to click through to read. Yworo (talk) 23:41, 6 July 2010 (UTC)

Yes, I've retracted my agreement, I did a mistake, I agreed for a mistake in understanding intentions, I've already said it's my fault. It's not like I've signed a blood pact. About "weaking the language", well, yes, I'd say that a full clumsy sentence focused only on the status of usability of GNU fails WP:UNDUE. Yet I am trying to compromise. "Interminable argumentation" is what I call "rational discussion", while you seem to behave as you owned the article. About the POV pushing, well, what POV am I pushing? Because I don't know, really, unless my "POV" is "I'd like to have a clear, neutral lead on the GNU article". --Cyclopia^talk 23:47, 6 July 2010 (UTC)

I notified the Wikipedia_talk:WikiProject_Software/Free_Software, in the hope someone can help in the discussion. --Cyclopia^talk 23:47, 6 July 2010 (UTC)

And don't wait for a response but revert to previous wording. Uh huh. Yworo (talk) 00:18, 7 July 2010 (UTC)

Notified also Wikiproject Linux and Wikiproject Computing. --Cyclopia^talk 00:54, 7 July 2010 (UTC)

I had a merge problem and unintentionally deleted the following comment. Sorry about that --Kvng (talk) 18:21, 8 July 2010 (UTC)

The arguing here is getting ridiculous. Work out the wording here, or in a subpage, and leave the page protected until there is agreement. Given that the definitions involved are somewhat nebulous, it really would make sense to drop the general statements from the lede, and have a section further down addressing the state and what people say about it - ironically, the argument over GNU here is a microcosm of a widespread debate, and therefore worthwhile for WP to report on in some depth (*without* taking sides). Stan (talk) 13:42, 7 July 2010 (UTC)

I would personally edit out all the information about what's not working with GNU. Who cares? The notable thing about GNU are all the working pieces (tools) that are in common use on many platforms. --Kvng (talk) 02:31, 7 July 2010 (UTC)

Thanks Kvng. Well, edit out all is a no-no. The current status of an OS is an important feature, especially in the case of GNU which is a quite notable failed promise in this respect and has been discussed in sources. I support inclusion of the status in the article lead, I only disagree on forcing it in the first sentence -in general, the lead as it is now is in a pretty bad state, and refocusing priorities of what to write there would help. --Cyclopia^talk 02:35, 7 July 2010 (UTC)

Kvng has a point. Why not redirect to GNU project. There is no GNU OS. Yworo (talk) 04:34, 7 July 2010 (UTC)

So much for NPOV. --Cyclopia^talk 09:53, 7 July 2010 (UTC)

I think this is all kind of lame, in a GF way. I don't like keeping this article protected. I'll be happy to unprotect if Cyclopia and Yworo agree not to edit the article at all until 10 July, not as a sanction of any kind, but only to give other editors a go at this. How 'bout it? Gwen Gale (talk) 09:02, 7 July 2010 (UTC)

If Yworo agrees, I agree. --Cyclopia^talk 09:49, 7 July 2010 (UTC)

I stopped editing after three attempts at compromise language close to my requirements. You're the one that kept going. Yes, Gwen, I'd stopped and didn't intend to continue. There was no need to protect the page. I'd also made that clear to Cyclopia before he asked for protection. Yworo (talk) 13:55, 7 July 2010 (UTC)

Yworo has caused this edit warring by constantly reverting edits that are not in accordance with his POV and not trying to reach consensus. In the references he uses to "support" his POV he is always using sources which represents the personal POV of the authors. They are not scientific or academic sources. For example (Steven Vaughan-Nichols "Opinion: The top 10 operating system stinkers" in Computerworld, April 9, 2009) It even includes the word "opinion". So it is not an objective source. The other "source" by Richard Hillesley reflects also the POV of the author and refers to the GNU Hurd. This article is about the GNU Operating system. The GNU Hurd has its own page. Lessig does not use any of the words pushed by Yworo. Lessig is referring to the kernel. GNU is a very flexible and modular operating system that works with different kernels. It does not mean that GNU is not complete. It is the Hurd that is not ready. If he wants to write about the Hurd then Yworo should go to the GNU Hurd page.--Grandscribe (talk) 10:26, 7 July 2010 (UTC)

Grandscribe, you clearly haven't bothered to catch up with the discussion. The sentence currently being discussed is "GNU is a computer operating system which has been under development since 1983 but which has not yet had its first stable release." There is nothing whatsover POV about it; it is completely accurate. I have been editing in perfectly good faith. You have made a personal attack and you need to make a retraction and an apology. Yworo (talk) 13:58, 7 July 2010 (UTC)

The arguing here is getting ridiculous. Work out the wording here, or in a subpage, and leave the page protected until there is agreement. Given that the definitions involved are somewhat nebulous, it really would make sense to drop the general statements from the lede, and have a section further down addressing the state and what people say about it - ironically, the argument over GNU here is a microcosm of a widespread debate, and therefore worthwhile for WP to report on in some depth (*without* taking sides). Stan (talk) 13:42, 7 July 2010 (UTC)

GNU Hurd has good NPOV. This article should be summarizing what's over there. Shouldn't be so hard. I agree with Gwen, there is a good faith problem here. Lame --Kvng (talk) 13:45, 7 July 2010 (UTC)

Yworo please stop accusing others of not acting in good faith as you have done against Cyclopedia. While we are trying to be objective and improve this article you persist on the inclusion of your POV on the very first sentence. The 3 references you use to support your POV are not valid since they express the "personal opinions" of the authors of each respective article. They are not indisputable scientific facts. The quality of this article is, unfortunately, suffering a lot since you started with your edits.--Grandscribe (talk) 09:57, 8 July 2010 (UTC)

Grandscribe, the sentence "GNU is a computer operating system which has been under development since 1983 but which has not yet had its first stable release." is clearly true and can be supported solely with citation to FSF documents. The opinions support the facts. Third-party opinions are always relevant because an NPOV article cannot be written about a product using only primary sources self-published by the organization which produces the product. Taking FSF claims at face value is not encyclopedic. Yworo (talk) 17:15, 8 July 2010 (UTC)

Third opinion

(disclaimer: I'm the Antichrist.)

I really don't think the recent changes have improved the lede. There is no requirement for the very first sentence to sum up the current working state of the system, precisely because it's too murky to ever be able to do so neutrally in a single sentence. All that is needed in the first paragraph is to say that while the project has been in development since 1983 a complete release of the OS has not occurred yet. Leave it to paragraph 3 to expand upon what "complete" means. Chris Cunningham (not at work) - talk 13:51, 7 July 2010 (UTC)

I agree with most of your comment. I object however using "complete/incomplete", a "stable/unstable" or "production ready/not ready" wording would be better. I also proposed an alternative lead above -your comments on that? --Cyclopia^talk 14:00, 7 July 2010 (UTC)

This article paragraphs should not include any kind of language that is not objective. Yworo has been the only one of all editors here pushing for words in the description of GNU that reflect his point of view. As support for his edits he can only give references to articles that represent the personal point of view and opinions of the authors and which are not indisputable scientific facts. --Grandscribe (talk) 09:08, 8 July 2010 (UTC)

The word "complete" is right from the GNU manifesto. Using the antonym is, I agree, somewhat troublesome, but that can be worked around better simply by describing what has actually been released (GNU 0.2, in 2002) rather than using euphemisms. Chris Cunningham (not at work) - talk 10:26, 8 July 2010 (UTC)

Sounds fine to me. --Cyclopia^talk 11:50, 8 July 2010 (UTC)

The word "complete" in the GNU manifesto is used in the context of future tense, specifically: "GNU, which stands for Gnu's Not Unix, is the name for the complete Unix-compatible software system which I am writing so that I can give it away free to everyone who can use it." Note the "which I am writing". This cannot be used to support either past or current completeness of the OS, because it is describing something yet to be written. The GNU manifesto was written in 1985 and it is clear that there was no "complete" operating system available at that time. There still is no complete OS, until Hurd is finished. Yworo (talk) 17:21, 8 July 2010 (UTC)

Rewriting the lead -restart

It seems to me that the majority of editors above (Grandscribe, Chris Cunningham and Kvng) agree that putting the unstable status of GNU in the lead sentence is not a priority.

Therefore I propose this wording for the lead, which is obviously open to criticism and comment. I suggest we begin working on the lead here and we discuss it here, before messing in the article again. I moved the stability sentences below because I feel that the aim of the OS is more important, in its encyclopedic description, than its current usability status, but it is of course open to change.

GNU is a free Unix-like operating system developed by the GNU project. It is composed wholly of free software: all the underlying source code can be used, freely modified, and redistributed, both commercially and non-commercially, by anyone under licenses such as the GNU General Public License. To paraphrase Richard Stallman, leader of the GNU project, the GNU system is a technical means to a social end. Despite aiming at releasing a "complete Unix-compatible software system" since its start in 1983, GNU has not yet issued a stable release which includes its official GNU Hurd kernel. GNU/Linux, which substitutes the Linux kernel to the GNU Hurd, is however generally considered production ready; also other free kernel replacements exist with varying degrees of stability.

--Cyclopia^talk 11:59, 8 July 2010 (UTC) (reworded a bit: --Cyclopia^talk 12:03, 8 July 2010 (UTC))

Let's put the status in the second sentence then. What you've proposed here is too convoluted, doesn't flow well at all and appears to imply that Linux could be considered an official part of GNU. The way we had it a couple of days ago was much less convoluted and much more accurate. To be clear, I mean this version:

GNU (pronounced /ˈɡnuː/) is a computer operating system. It has been under development since 1983 but has not yet had its first stable release. However, production use is possible using a Linux kernel or other free kernel. GNU is composed wholly of free software. Its name is a recursive acronym for “GNU's Not Unix!” This name was chosen because GNU's design is Unix-like, but differs from Unix by being free software and containing no Unix code. Development of GNU was initiated by Richard Stallman and was the original focus of the Free Software Foundation (FSF).

-- Yworo (talk) 17:10, 8 July 2010 (UTC)

Both versions have their merits. A compound sentence or two is good for flow. Somewhere there should be a reference to the GNU Project as the moving force, since that is technically different from the FSF, and the "not yet issued" needs an "as of 2010" or some such, because maybe the system will see its first stable release next year. :-) Stan (talk) 17:19, 8 July 2010 (UTC)

I do think putting the start date and current status in a single sentence is good. Keeping them together and in chronological order helps the flow. Moving from the status of the official GNU with Hurd directly to its usability with a Linux kernel or other kernel is also good flow. We should perhaps move the "composed wholly of free software" into the first sentence. I don't think we should add details about what free software is to the lead paragraph, as free software is wikilinked. Certainly the additional information about what free software is could go into a subsequent paragraph of the lead. Mentioning the GNU project in the lead paragraph as you suggest certainly seems apropos to me. Perhaps the explanation of the name should be deferred to the next paragraph or even put into a footnote. Yworo (talk) 17:31, 8 July 2010 (UTC)

Cyclopia's proposed text is a very good and objective attempt to find a solution to use as a description of GNU. I am in favor of it instead of any other version that uses subjective language (unstable, incomplete, etc) in the first paragraph. As some well meaning editors, who have taken part in the discussion, suggest the opinions could be placed, if at all appropriate according to wikipedia standards, in a separate paragraph at the end of the article and clearly identified as "opinions".--Grandscribe (talk) 21:19, 8 July 2010 (UTC)

Neither current proposal puts "unstable" or ""incomplete" in the lead. My current proposed simply highlights the fact that no release designated stable has yet been made. Certainly version 0.2 is not designated "stable". And Debian GNU/Hurd, which is technically distinct from GNU, gives it's status as "not officially released yet, and won't be for some time".[5] Certainly a bit more substantial than vaporware, perhaps it could be called "steamware"? Although that suggests steampunk, so maybe not. :-) Yworo (talk) 21:57, 8 July 2010 (UTC)

Proposal by Cyclopia

Ok, let's see if I can merge some of the suggestions above in an updated proposal:

GNU (pronounced /ˈɡnuː/)is a free Unix-like operating system developed by the GNU project. It has been under development since 1983 but has not yet had its first official stable release. GNU/Linux operating systems, which substitute the Linux kernel to the official GNU Hurd, are however generally considered production ready; also other free kernel replacements exist with varying degrees of stability. GNU is composed entirely of free software. Its name is a recursive acronym for “GNU's Not Unix!” , chosen because GNU's design is Unix-like, but differs from Unix by being free software and containing no Unix code. Development of GNU was initiated by Richard Stallman and was the original focus of the Free Software Foundation (FSF). Richard Stallman, leader of the GNU project, declared that the GNU system is a technical means to a social end.

--Cyclopia^talk 22:59, 8 July 2010 (UTC)

I'm concerned about the use of "GNU/Linux" there outside of quotation marks. This is the FSF's branding; it's not a neutral description of the resulting OS, as discussed to death on talk:Linux. The "technical means to a social end" bit is fluff which could probably be safely omitted. In general, I think the flow of your first draft was superior. Chris Cunningham (not at work) - talk 08:51, 9 July 2010 (UTC)

I have no strong feelings on the infamous naming controversy, but in this specific case to indicate only Linux could easily mislead people into thinking that Linux is a part of GNU. What about "Variants which substitute the Linux kernel to the official GNU Hurd kernel, sometimes called GNU/Linux..." ? About the "technical means to a social end", well, it is the very aim of GNU's existence, not random fluff, whatever one thinks of it. --Cyclopia^talk 13:53, 9 July 2010 (UTC)

The whole "Linux is a kernel, GNU is an OS" thing is largely the FSF's branding as well. That's what makes this whole thing so troublesome. We need to avoid presenting the subject from the perspective of the FSF unless we're making it clear whose position it is. As for the trailing text, I think that it is obvious enough already from the rest of the text that GNU is the implementation of the FSF's main goal of a completely free OS. Chris Cunningham (not at work) - talk 14:35, 9 July 2010 (UTC)

Neutral comment. I may have stirred that up with this post on my talk page (which indeed notes the controversey). The pith of my thinking is, a helpful lead would quickly and thoroughly bring readers up to speed as to what the topic is all about, even saying there is controversy and calling it such, rather than hew to some notion of "consensus PoV" one way or another. Gwen Gale (talk) 09:12, 9 July 2010 (UTC)

As for technical means to a social end, doesn't matter a mote if it's Stallman channeling the canny pith of fate, liberty and happiness through helpful tool making, or muddled marketing codswallop stained with clueless notions of utopian social engineering scams, or whatever else, it's notable. Moreover, sourced opinions as to what it means and what folks think about that are welcome. Whether it belongs in the lead may have a lot to do with how it's handled in the lead. Gwen Gale (talk) 14:17, 9 July 2010 (UTC)

I'd be happy with this version as long at that third sentence could be rewritten in a way that does not use "GNU/Linux", which is POV. Yworo (talk) 14:24, 9 July 2010 (UTC)

What about a wording of the kind ""Variants which substitute the Linux kernel to the official GNU Hurd kernel, called usually Linux and sometimes GNU/Linux..."? I'd say we could keep both namings. --Cyclopia^talk 14:29, 9 July 2010 (UTC)

Or, to address Chris Cunningham further POV worries, "Variants which substitute the Linux kernel to the official GNU Hurd kernel, called usually Linux (or GNU/Linux in the FSF nomenclature)..."

How about GNU utilities are frequently combined with a Linux kernel rather than GNU Hurd to produce Linux, a production-ready operating system also referred to as "GNU/Linux";... Yworo (talk) 14:44, 9 July 2010 (UTC)

The problem is that "Linux is a variant of GNU" is a dictinctly minority position, held primarily by the FSF and its affiliates. It's very tricky to word the situation without either obviously contradicting either the FSF (which maintains that GNU is a widely deployed operating system) or Linus Torvalds (who made a distinction between Linux-the-OS and GNU-the-OS on day 1). What about just saying "the FSF maintains that Linux, when used with GNU tools and utilities, should be considered a variant of GNU, and promotes the term GNU/Linux for such systems"? That way we're not making that call, but simply pointing out what their position is. Chris Cunningham (not at work) - talk 14:52, 9 July 2010 (UTC)

I'd say it is an excellent wording. --Cyclopia^talk 15:01, 9 July 2010 (UTC)

I also agree that making it clear that this is an FSF position is important. Yworo (talk) 15:04, 9 July 2010 (UTC)

New iteration. Probably it doesn't flow as well as some of the others but it tries to address concerns...

GNU is a Unix-like computer operating system developed by the GNU project, ultimately aiming to be a "complete Unix-compatible software system" composed wholly of free software. Development of GNU was initiated by Richard Stallman in 1983 and was the original focus of the Free Software Foundation (FSF), but no stable release of GNU yet exists as of July 2010. The latest alpha release of the GNU system is GNU 0.2, released in 2004, featuring GNU Hurd as the system's kernel. Other (non-GNU) kernels can also presently be used with GNU; the FSF maintains that Linux, when used with GNU tools and utilities, should be considered a variant of GNU, and promotes the term GNU/Linux for such systems (leading to the GNU/Linux naming controversy). Its name is a recursive acronym for “GNU's Not Unix!” , chosen because GNU's design is Unix-like, but differs from Unix by being free software and containing no Unix code.

--Cyclopia^talk 15:57, 9 July 2010 (UTC)

Yes, you are right, it doesn't flow as well, but content and content ordering are fine. If we must have discussion of what free software is, it can go in the second (or later) paragraph of the lead. The first paragraph needs to focus on GNU, what it is, who created it, the actual status and the reasons some disagree with that status due to variants. Yworo (talk) 16:03, 9 July 2010 (UTC)

Oh, "still" should be removed or replaced with "yet". Yworo (talk) 16:04, 9 July 2010 (UTC)

 Done

OK! Chris, others, any comment? If no concerns on content organization, we could put this up for a start (with a big hidden comment saying "if you want to touch it, please go to talk page first", given the pain it took) and then finalize the flow? --Cyclopia^talk 16:07, 9 July 2010 (UTC)

Whoops, missed this. I rather prefer the iteration in the section below. Chris Cunningham (not at work) - talk 10:15, 11 July 2010 (UTC)

Proposal by Yworo, based on Cyclopia's with Chris Cunningham's suggestion

Here's a modification of Cyclopia's proposal with several suggestions applied:

GNU (pronounced /ˈɡnuː/) is a Unix-like operating system, developed by the GNU project, which is composed entirely of free software. It has been under development since 1983 but as of July 2010 has not yet had its first official stable release. The FSF maintains that Linux, when used with GNU tools and utilities, should be considered a variant of GNU, and promotes the term GNU/Linux for such systems; also other free kernel replacements exist with varying degrees of stability. Its name is a recursive acronym for “GNU's Not Unix!” , chosen because GNU's design is Unix-like, but differs from Unix by being free software and containing no Unix code. Development of GNU was initiated by Richard Stallman and was the original focus of the Free Software Foundation (FSF).

(added as of) Yworo (talk) 15:37, 9 July 2010 (UTC)

I tried a new iteration above based on yours and Chris latest versions. --Cyclopia^talk 15:59, 9 July 2010 (UTC)

Proposal by Chris Cunningham

Okay then: let's see...

GNU is a Unix-like computer operating system developed by the GNU project under the auspices of the Free Software Foundation (FSF). It is composed wholly of free software: all the underlying source code can be used, freely modified, and redistributed, both commercially and non-commercially, by anyone under licenses such as the GNU General Public License.

GNU was conceived in 1983, when Richard Stallman, the GNU project's founder, raised a call for a "complete Unix-compatible software system". The latest release of the GNU system is GNU 0.2, released in 2004, featuring GNU Hurd as the system's kernel (this is also referred to as GNU/Hurd). Other kernels can also presently be used with GNU; the FSF maintains that Linux, when used with GNU tools and utilities, should be considered a variant of GNU, and promotes the term GNU/Linux for such systems (leading to the GNU/Linux naming controversy).

How's that? Needs some expansion regarding design and implementation, and possibly reception, but I think that sorts out the hard bits. Chris Cunningham (not at work) - talk 15:20, 9 July 2010 (UTC)

Uh, no. I prefer Cyclopia's version with the third sentence changed as discussed. Yworo (talk) 15:23, 9 July 2010 (UTC)

(edit conflict) It is very good, but some mention must be made of the fact that no stable version of the official GNU exists -this was roughly consensual above (even if there was strong disagreement on how to include it): what about this edit?

GNU is a Unix-like computer operating system developed by the GNU project under the auspices of the Free Software Foundation (FSF). It is composed wholly of free software: all the underlying source code can be used, freely modified, and redistributed, both commercially and non-commercially, by anyone under licenses such as the GNU General Public License.

GNU was conceived in 1983, when Richard Stallman, the GNU project's founder, raised a call for a "complete Unix-compatible software system". However, no stable release of GNU still exists as of 2010, the latest alpha release of the GNU system is GNU 0.2, released in 2004, featuring GNU Hurd as the system's kernel (this is also referred to as GNU/Hurd). Other kernels can also presently be used with GNU; the FSF maintains that Linux, when used with GNU tools and utilities, should be considered a variant of GNU, and promotes the term GNU/Linux for such systems (leading to the GNU/Linux naming controversy).

--Cyclopia^talk 15:27, 9 July 2010 (UTC)

The defect of this one is the explanation of free software in the first paragraph. This article isn't the article on free software, it the article on the GNU operating system. We have a prefectly good article on free software and don't need to explain it again here. Please see my alternate proposal which integrates various suggestions just above. Yworo (talk) 15:35, 9 July 2010 (UTC)

Articles should be able to stand alone to a reasonable extent. As one of the most high-profile articles we have on free software (the subject of the manifesto which kicked it off, no less), it's not unreasonable to explain what free software is in the lede without expecting the reader to go reading up on free software for context. Chris Cunningham (not at work) - talk 09:22, 11 July 2010 (UTC)

Sure, but certainly the first paragraph should be about the subject of the article! Yworo (talk) 15:03, 11 July 2010 (UTC)

I'm very happy with this one; if there are no further objections I think we can roll this one out. Chris Cunningham (not at work) - talk 09:23, 11 July 2010 (UTC)

I'm not. The first paragraph needs to focus on GNU. If you want to further explain free software in the second paragraph, I have no objections to that. Yworo (talk) 15:04, 11 July 2010 (UTC)

Agree with Yworo on that, better to explain free software later (I've put it there first , before, in analogy with Linux lead, but it isn't a good choice). I'd say we could try to deploy the "15:57, 9 July 2010 (UTC)" iteration above (which, even if not perfect at all, seems to be a step in the consensual direction) and then continue tweaking on the talk page. Chris, could you explain why you prefer this version to the other one? --Cyclopia^talk 15:36, 11 July 2010 (UTC)

For the reason given previously: GNU-the-OS is the primary focus of the GNU project and one of the highest-profile free softwware projects. Linux actually has a longer explanation of the same thing in the first paragraph. But I'll see if I can work on a compromise in the article itself. Chris Cunningham (not at work) - talk 11:00, 13 July 2010 (UTC)

Not a good enough reason for first paragraph. In fact, it's a problem in the Linux article also. Yworo (talk) 12:36, 13 July 2010 (UTC)

The Linux lede was discussed to death over a period of years by rather more than three editors, and subsequently passed GA. I would far rather take its cue. Chris Cunningham (not at work) - talk 23:41, 13 July 2010 (UTC)

I've seen Yworo has reworked the Linux lead paragraph and it reads indeed better now. Anyway, in the case of Linux it could have some sense in being possibly the most prominent example of free software. In the case of GNU I am not so sure. --Cyclopia^talk 10:37, 14 July 2010 (UTC)

Proposal by Grandscribe

What about this draft? I made some changes to your proposed version:

GNU (pronounced /ˈɡnuː/ is a free Unix-like computer operating system developed by the GNU project. Its development started in 1983. An official release using the project's own kernel, the "GNU Hurd" has not been distributed yet. GNU/Linux operating systems, which substitute the Linux kernel to the official GNU Hurd, are however generally considered production ready; also other free kernel replacements exist with varying degrees of stability. GNU is composed entirely of free software. Its name is a recursive acronym for “GNU's Not Unix!” , chosen because GNU's design is Unix-like, but differs from Unix by being free software and containing no Unix code. Development of GNU was initiated by Richard Stallman and was the original focus of the Free Software Foundation (FSF). Richard Stallman, leader of the GNU project, declared that the GNU system is a technical means to a social end.

--Grandscribe (talk) 06:57, 9 July 2010 (UTC)

It doesn't flow nearly as well as Cyclopia's proposal. Yworo (talk) 14:27, 9 July 2010 (UTC)

Rolled out

I've requested unprotection as we seem to be getting somewhere, and rolled out Cyclopeia's proposal from 15:57, 9 July 2010 as suggested above. There's some material in the old lede which isn't covered by the new one, so there's more work to do here, and I'd rather incorporate some of the edits in Cyclopeia's other revision; but we can work on that in the article itself I think. Chris Cunningham (not at work) - talk 12:03, 13 July 2010 (UTC)

Excellent, thank you! --Cyclopia^talk 12:26, 13 July 2010 (UTC)

References

Adding a link

I'd like to make the word GNU in this sentence: "GNU's Not Unix!" a link to the same page! :D it reveals the recursion of it even more! —Preceding unsigned comment added by Nima101 (talk • contribs) 17:30, 8 July 2010 (UTC)

It's pointless and against our policies. Pages should not self-link to themselves. In fact, doing so would not create a link, but would simply bold GNU, because the Wikimedia software is smart enough not to bother. Yworo (talk) 17:32, 8 July 2010 (UTC)