Talk:Linux/Archive 18
![]() | This is an archive of past discussions. Do not edit the contents of this page. If you wish to start a new discussion or revive an old one, please do so on the current talk page. |
Archive 15 | Archive 16 | Archive 17 | Archive 18 | Archive 19 | Archive 20 | → | Archive 25 |
Info box only for the kernel?
Hi, when I changed the info box at the top of the page to reflect the fact that the (GNU/)Linux operating system contains software under licences other than the GNU GPL, it was changed back, with the reason given being that the info box is just for the kernel, rather than the whole operating system. Why is that? Surely as the article is about the whole system, the information in that box should be as well? Guyjohnston 19:41, 8 May 2007 (UTC)
- Infoboxes are meant to be at-a-glance guides. The GPL is the lowest common denominator for most distros. Chris Cunningham 21:02, 8 May 2007 (UTC)
- That's fair enough that we don't want it to be too confusing. However, I've added "and others" to be more accurate by showing that the system contains some non-GPLed software, without adding a lot of confusion. Guyjohnston 23:14, 11 May 2007 (UTC)
- Agreed, most linux distros contain software that is under non gpl compatible free software licenses, generally variants of 4 clause BSD, also MPL and variants (though the main mozilla code has now been relicensed bugzilla and various other projects using MPL or MPL like licenses have not). Plugwash 14:32, 18 May 2007 (UTC)
- It would be best if this discussion took place on Template talk:Linux. Gronky 14:39, 18 May 2007 (UTC)
Recent edits
- I don't understand the rewording of sentence two. It needlessly pipes over links to articles which exist (free and open source software) and loses the distinction the sentence currently has.
- The rewording of the kernel licence change also loses the whole point of changing the licence, with an edit summary which suggests that it's misleading. It isn't misleading. The intricacies of what constitutes a derivative work are unimportant here, really. The new ref is also less useful, as it presents the GNU message from a primary source instead of being an interview on the actual subject.
- The philosophy edits are fine.
- The moving of GNU/Linux back into the first line is contentious. This is a minority position and is already adequately addressed in the intro.
I don't see that these edits did much to improve understanding of the article.
Chris Cunningham 21:02, 8 May 2007 (UTC)
- Fair enough. I think that part's fine either way
- I was correcting an inaccuracy. The previous wording implies that only software licensed under the GPL can be included in the (GNU/)Linux system. That's simply not true. The evidence for that is that all distributions (at least the major ones) contain software under other licences, such as the LGPL, and most of them also contain some proprietary software. According to the previous wording all of those distributers are violating the GPL by doing that. I've now changed that back. The original reference is probably better. I didn't read it very well before so I thought it was just about not switching to the GPL v3.
- OK
- Fair enough if that's been discussed and agreed on. I had a quick look in the discussion page before and didn't see that.
- Guyjohnston 23:27, 11 May 2007 (UTC)
Philosophy section
I don't see any evidence that the primary difference between Linux and other contemporary systems is that Linux is free software. After all, there are other contemporary free-software OSes that differ remarkably from Linux, such as the BSDs -- or to go further afoot, Atheos or Haiku. Likewise there are other systems that are Unix-like but not free, such as Mac OS X.
It is original research to call free software the primary difference between Linux and other systems. --FOo 04:41, 9 May 2007 (UTC)
- "other popular contemporary operating systems" I'm sure you don't claim that Atheos or Haiku are popular (even BSD is kind of dubious in that category. This paragraph compares Linux with Windows and possible Mac OS x (which is not pure BSD and has proprietary stuff AFAIK) not with Atheos and Haiku... and I think it's a pretty defining difference. That being said I'm fine with "one of the difference", maybe it would be better "one important difference" but that's a POV (but I'm sure we can find enough references for it) -- AdrianTM 05:55, 9 May 2007 (UTC)
- Fair enough. --FOo 06:39, 9 May 2007 (UTC)
Heavy reordering
This has been tried before and met with resistance (mostly due to conflicting ideas about where the article should be heading), but I think it's time to give it another go. The last half still needs a lot of work on direction (including killing a lot of editorialising), but I think the general ordering is better this way (moving more important things up and more impact-based things to the end). Chris Cunningham 09:24, 12 May 2007 (UTC)
Criticism article
Well it has been a few weeks now since the article Criticism of Linux had its AFD. It was kept but then majourly tidied/cut down. As it stands, the article consists of 2 sections covering the viability of Linux on the desktop and covering TCO and Microsoft campaigns. These 2 sections are appropriate as information to add to either the 'Desktop' section of this article or the Desktop linux article and the 'Market share and uptake' section or the Linux adoption article. I recommend that we merge the article into either this page or its sub articles as it is not likely going to grow much more than it is - due to a major lack of up-to date reliable sources. What do others think? My other reason for the merge is that the criticism article at present mixes up the 'desktop' and 'server' criticisms quite badly - it isn't obvious what things are referring to, putting it in their relevant sections/sub-articles would solve this.-Localzuk(talk) 16:34, 12 May 2007 (UTC)
- I agree as long as the info is added into this article where it belongs, not in a separate "criticism" troll/POV magnet section. - AdrianTM 20:10, 12 May 2007 (UTC)
- That might be a good idea. However, the article is already quite long, so I don't think it should be added to unless it's seen as completely necessary. I've noticed that a warning appears when you edit the page which reads "This page is 53 kilobytes (kibibytes) long. It may be appropriate to split this article into smaller, more specific articles. See Wikipedia:Article size." Guyjohnston 14:37, 13 May 2007 (UTC)
- Yes, the article is long already, but having a very small criticism article is pointless. That is why I indicated we should maybe merge it into the existing relevant subarticles of this article. It just doesn't stand very well on its own.-Localzuk(talk) 14:42, 13 May 2007 (UTC)
- That might be a good idea. However, the article is already quite long, so I don't think it should be added to unless it's seen as completely necessary. I've noticed that a warning appears when you edit the page which reads "This page is 53 kilobytes (kibibytes) long. It may be appropriate to split this article into smaller, more specific articles. See Wikipedia:Article size." Guyjohnston 14:37, 13 May 2007 (UTC)
- I am a bit reluctant to add a "criticism" section to Linux. We have a full article describing Linux, and another one comparing it to Windows. The reader should be able to conclude what is good/bad in Linux, without being told so, shouldn't she? I mean, we do not have an article for the Merits of Linux, do we? Specific "Criticism of..." sections are OK if they point out obscure but important facts, not readily understandable from the bare descriptions. But as the Criticism of Linux article stands now, it consists of a) a mild rant of an ex Windows user complaining that she misses the plethora of software she could download from eMule for her Windows machine (outright false, because there are more software alternatives in Linux than in any other platform, except for specific commercial programs), and that her modem won't work in Linux (which is a problem of the modem), and b) a brief parroting of Microsoft's obviously biased pamphlet. The first paragraph could be adapted and included in Linux... maybe. The second, no way. — Isilanes 17:52, 15 May 2007 (UTC)
- I agree with Localzuk, and to avoid the 53kb warning we could remove a few pointless facts and infomation from both articles - Solidsnake204 07:33, 16 May 2007 (UTC)
- I will make a start tomorrow, as people will have had 7 days to comment then. I will be going for the 'put the criticism where it is appropriate' method - so there will be a brief summary of it in this article and the bulk will go in the other subarticles I reckon.-Localzuk(talk) 16:55, 18 May 2007 (UTC)
- I agree with Localzuk, and to avoid the 53kb warning we could remove a few pointless facts and infomation from both articles - Solidsnake204 07:33, 16 May 2007 (UTC)
- In the timeline of unix it says GNU/Linux <linux version>. since this tries to be only linux and not gnu, using gnu word incorrect like this would seem even worse. when did Linux come part of the GNU ptoject ;)
- How can you not have a criticism of Linux page when you have a similar page for just about every release of Windows? There are valid criticisms dealing with Total Cost of Ownership, ease of use, keeping up with all the different versions, lack of standard documentation and reliance on the command line.
History ordering
This was changed a while back, primarily because while GNU technically predates Linux this isn't actually an article on GNU. What's the current consensus on this? Chris Cunningham 22:26, 20 May 2007 (UTC)
- It's true that this isn't an article about GNU, but it's not an article about Linux, the kernel, either. It's an article about the complete operating system made from combining some of the GNU components with Linux and other software. Therefore, I think the history section about when the different parts were written should be in chronological order. Guyjohnston 14:27, 21 May 2007 (UTC)
- Cool. In that case I think we should make some effort to provide a very brief summary of what people were doing writing Unices in the first place. I've added some placeholder text. Chris Cunningham 18:11, 25 May 2007 (UTC)
- Yep sounds like a good idea. I seem to remember reading that X and Latex were two pieces of free software which were available and useful to the GNU developers when they started, so it might be worth mentioning them near the beginning. Guyjohnston 01:11, 29 May 2007 (UTC)
Release date of the OS?
In this article, the names "Linux" and "GNU/Linux" refer to an operating system, so if a sentence says "Linux was released on DateX", then "DateX" must be the release date of the operating system. Someone added such a sentence to the intro, and for "DateX" they used "September 17th 1991".
This is patent nonsense since the operating system didn't exist on that date. The kernel-less GNU operating system and the v0.01 of the Linux kernel existed, but that version of the Linux kernel was not functional enough to support an operating system, and even when the Linux kernel became more feature-complete, there was still the work of combining it and GNU to make a complete operating system.
As a secondary point, there is probably no day, or even month, that can be claimed as the release date of the operating system because it's transition from not existing to existing was a gradual evolution with no one key date, and also because for it to be released, someone would have to have released it, but I don't think anyone agrees, or even claims, that there is one person or organisation that definitively released the operating system.
Clearly, someone has gotten confused by the use of the kernel's name to refer to the operating system. For clearer thinking, should the article say this: "GNU/Linux was released on September 17th 1991"? Of course not.
That's why I've removed the sentence "Linux was released on Sep 17 1991" from the intro. At best, the intro could give the year that the operating system became usable. That was probably 1992, but I'm just guessing. For the moment I've left it as 1991 in the intro because that's the year that the previous person used. Can someone give any insight into which year is correct? Gronky 14:27, 23 May 2007 (UTC)
- I think we need to use the date of the first release of a "Linux distribution" (no matter what feature had or what featured it missed). -- AdrianTM 14:54, 23 May 2007 (UTC)
- We've discussed "completed" before. This is a POV statement from the GNU camp. Linux is arguably still not "complete", so it doesn't make sense to pick any arbitrary point except for the first release of the kernel (which is when there was released code called "Linux"). And again, I'd advise people to keep in mind that this is not "the GNU/Linux article but called Linux"; it's named the way it is because of a generally blurriness in the use of the term "Linux", and the kernel has an article to itself for style reasons more than because it's a distinct entity from "the operating system". Chris Cunningham 16:21, 24 May 2007 (UTC)
- I agree with Adrian. The only possible date we could use is the initial release date of the first distribution. But I'd like to formulate it that way: "the first Linux distribution was released 1234" and not "Linux was released 1234". Of course I by Linux I meant GNU/Linux really. --MarSch 10:17, 25 May 2007 (UTC)
- This article isn't about "GNU/Linux but just called Linux". GNU/Linux is an invented term which seeks to draw an arbitrary border between operating systems and kernels where there may be no strict line (such as in Linux) for the sake of advertising. Even the term "first distribution" isn't clear, given the unclear definition of exactly what constitutes a distro.
- First distribution of source is a perfectly valid way of dating initial release. In this case, I'd argue that it's the only one which can be stated with certainty. Chris Cunningham 11:16, 25 May 2007 (UTC)
- I agree with MarSch. I think that the term "Linux distribution" should be preferred in this case. -- AdrianTM 14:36, 25 May 2007 (UTC)
- Why pick an arbitrary package instead of going with the initial code drop? Distros are considerably less "official" than kernel releases, at any rate. I'm yet to be convinced on this line of reasoning. Chris Cunningham 18:05, 25 May 2007 (UTC)
- It's because we don't talk about Linux kernel, we talk about Linux OS here. -- AdrianTM 18:27, 25 May 2007 (UTC)
- This is an arbitrary line. Torvalds himself refuses to see it. When you have graphics card drivers in the kernel, doesn't that mean X is part of the kernel? These are murky areas and should be treated as such. Chris Cunningham 20:56, 25 May 2007 (UTC)
- I don't see any arbitrary line there is the kernel and there is anything else beyond the kernel, there's no possibility of confusion between things. It's one thing if we talk about Linux kernel and another if we talk about the system composed from kernel, tools, X, KDE, etc. To see that's a clear demarcation just think that the kernel is developed by an entity and the rest is developed by other people/companies and can have totally different licenses. Linux kernel initially used non-free Minix tools for example. -- AdrianTM 21:15, 25 May 2007 (UTC)
- Errr, I just gave you a clear example of an instance where this is contradicted: graphics card drivers in the kernel. Then there's file systems, which are a total kernel thing in Linux but not considered to be so in some microkernel archs. There is no clear demarcation. There are a few murky ones. So the article should treat the difference as murky. Chris Cunningham 10:19, 26 May 2007 (UTC)
- What are you talking about? A graphic card that's in Linux kernel developed by Linus & Co. is part of Linux kernel by definition, a graphic card driver that's not included in Linux kernel at http://kernel.org/ it's not, where's the murkiness? Beside that's a minor and marginal issue: drivers, we talk here in this page about major things: X, KDE, GNOME that I doubt anyone can say that they are part of kernel in any way, especially that many are included in other OSes: BSD for example. -- AdrianTM 12:22, 26 May 2007 (UTC)
- That doesn't make much sense. Some drivers are part of the kernel, some aren't? The dividing line between Linux kernel and Linux is where a project's CVS is hosted? There's plenty of nuance. This isn't a clear-cut issue and should not be presented as such in the article by making out that there's a normative line between Linux-the-OS and Linux-the-kernel. Chris Cunningham 15:05, 26 May 2007 (UTC)
- I am not sure I understand your logic, how can you miss the clear cut, look at kernel.org there's the official Linux kernel in any other part if you have that kernel + some other stuff = distribution, some distributions are more complete then others, we need to decide if we talk about distributions or about kernel, since I see info about X, KDE, GNOME and other stuff I can only assume that we talk about distributions, the kernel is not even an entire OS from some points of view, it needs GNU stuff to qualify. By definition GNU stuff is not part of kernel, there's the clear cut, when somebody takes the kernel and add the rest and distribute then you get a "Linux distribution". I'm no GNU POV-pushing fan here, but we need to decide what we talk about here, a kernel is not an OS (unless in minor cases that are mostly ignored in this article) since we talk about OS here (presumably) then it makes sense to talk about the first distribution whichever that might have been. -- AdrianTM 06:04, 29 May 2007 (UTC)
- Sorry? X, KDE and Gnome are not part of the Linux OS. They can be part of the OS. In embedded systems or on the full multiuser/multitask systems (both *ix and Linux) that i used in the early '90s (and for some customers up to now) - there is no such thing. And its still the Linux OS. I agree that distribution is probably the WP:COMMONNAME usage for Linux btw. but distribution should not be set as a final word - as roll-your-own is still Linux and very much in use (in the embedded world). --Kim D. Petersen 19:46, 26 May 2007 (UTC)
- Those were examples for different licenses used, if you use this logic only the Linux kernel is only necessary think to be in a "Linux OS", which in a sense is true but that's not what we talk about in this page, here it seems like we talk about "Linux distributions" and the distributions have a different release schedule than the kernel I think it makes sense to use the first Linux distribution release date here and the kernel release date in kernel article (since these articles seem to be about different things, right? If they are about the same thing they should be merged) -- AdrianTM 06:04, 29 May 2007 (UTC)
- I disagree - i'd define Linux (as opposed to the kernel) as the kernel + at least a limited set of basic facilities: init, mount, shell, and so on. Basically what you find in something like Busybox. Of course one could just strap a program onto the kernel as if it was init - but thats going to far. --Kim D. Petersen 06:56, 29 May 2007 (UTC)
- Those were examples for different licenses used, if you use this logic only the Linux kernel is only necessary think to be in a "Linux OS", which in a sense is true but that's not what we talk about in this page, here it seems like we talk about "Linux distributions" and the distributions have a different release schedule than the kernel I think it makes sense to use the first Linux distribution release date here and the kernel release date in kernel article (since these articles seem to be about different things, right? If they are about the same thing they should be merged) -- AdrianTM 06:04, 29 May 2007 (UTC)
- The fact that there is no obvious initial release date for the "Linux operating system" shows that there is no such thing as "the Linux operating system". The only thing that can legitimately be called "Linux" without any additional qualifiers is the kernel. As I have suggested before, the contents of this article should be merged into Linux distribution and/or Linux kernel, and the article at Linux should be a redirect or a disambiguation page. — A.M. 02:35, 27 May 2007 (UTC)
- "Can be"? Can you point to a source which is qualified to give normative definitions? If not, this argument fails the same way as it's done previously. The most common usage of the word "Linux" in the wild refers to a complete OS. See the archives. Chris Cunningham 16:17, 28 May 2007 (UTC)
- That's not quite true: "Linux" in the wild refers to "Linux distribution". That's why A.M. proposal makes sense and I support it. Of course "Linux" should be redirected to "Linux distribution" not to "Linux kernel" because of the usage of the word, here probably A.M. and I part ways, but nevertheless "Linux OS" is an invented thing (and POVish depending where you draw the line), people when they talk about Linux they either talk about Linux kernel or a specific Linux distribution (or generic Linux distribution: kernel + nice suff). -- AdrianTM 06:04, 29 May 2007 (UTC)
- "The most common usage of the word "Linux" in the wild refers to a complete OS. See the archives." If so, then what is the release date of that complete OS? There's no ambiguity about the release date of Windows XP, nor about that of OpenBSD 3.7, so why is there ambiguity about the release date of "Linux"? Also, Wikipedia does not and should not necessarily use the most frequently-used name for something, especially when there is a less ambiguous name; for example, film vs. movie, Internet media type vs. MIME type, steel-string acoustic guitar or classical guitar vs. acoustic guitar, bicycle or motorcycle vs. bike, etc.
- "Of course "Linux" should be redirected to "Linux distribution" not to "Linux kernel" because of the usage of the word, here probably A.M. and I part ways" I don't object to Linux redirecting to Linux kernel. However, I would prefer the latter article to be called Linux (kernel). — A.M. 08:40, 29 May 2007 (UTC)
- First, our naming conventions do call for us to use the most common name - as has been said by dozens of editors dozens of times in our archives. Linux is the most common name for this OS plain and simple. There is no obvious 'release date' as the entire OS is a fluid set of projects that do an awful lot of releases at all sorts of times. They differ from commercial OS's such as Windows XP in that there are new versions coming out every day. We could easily pinpoint the first commercial release of a 'packaged' distro but pinning down the first release date of the actual OS is not possible.-Localzuk(talk) 11:02, 29 May 2007 (UTC)
Controversy stuff in the intro
Seriously, do we need to go into so much detail about the name in the intro? It's only two short sections until we get a decent elaboration on it. I'd prefer it if this article had as little duplication as possible and have worked to that effect so far. Chris Cunningham 18:05, 25 May 2007 (UTC)
- I don't understand why you object against an Wikipedia link that is highly pertinent to the naming issue. -- AdrianTM 18:26, 25 May 2007 (UTC)
- It's duplicate information which is of little importance to those readers who don't have to read FSF press statements very often. It's explained in detail, with a link, only slightly further down the article. Chris Cunningham 20:56, 25 May 2007 (UTC)
IP stuff (copyright, trademark, naming controversy etc)
I think this stuff really belongs in its own section:
- Right now it's in three or four places. There's some duplication in licensing because of this.
- It allows it to be presented as a complete section with subheaders. Right now because of the splitting, it's unevenly-weighted.
- It really forms the backbone of the majority of substantial (i.e. referenced) criticism of Linux, which gives the article more of an appearance of objectivity than having a section called "criticism" which is ripe for random trolling.
The issue is what to call it. "Intellectual property" is going to cause some free software people to want to chew tinfoil. I personally can't think of a better term for that abstract collection of legal stuff surrounding code, but thought I'd try asking. Chris Cunningham 18:19, 25 May 2007 (UTC)
- Thats a hard one. Wikipedia has to appeal to the masses not RMS so I think "Intellectual property" is an OK term for all of Copyright, Patents, DRM, Licensing etc. So far the "Intellectual property" regime of Linux is working very well with no substantiated claims of copyright violation, no substantiated claims of patent violation and no challenges to the licensing regime (i.e. GPL) even though there has been much criticism by competitors but none of the mud has stuck. In the highly litigious world of software and technology this is an excellent record. (Slackware -> Rehdhat -> Mandrake/Mandriva -> Ubuntu user and very small part-time Open Source coder :) Ttiotsw 12:46, 26 May 2007 (UTC)
Adding self-hosting date to intro ?.
In the world of compilers and computers self-hosting is the biggest hurdle in maturity after you have a clean compile. It is an especially important issue with Free (as in Freedom) software as it means you are not dependant on other software (a distinction that closed-source software based systems never really care about). From the Linux (kernel) article it says,
- "In December 1991, Linux 0.11 was released. This version was the first to be self-hosted - Linux 0.11 could be compiled by a computer running Linux 0.11."
Given it used the GNU tools too (which could be compiled using GNU tools + Linux 0.11) this is probably the first date that you could say that the Linux-as-a-computer-OS existed as a separate entity i.e. the date it cut the umbilical from Minix. OK to add (the bit in quotes indented above) ?. Ttiotsw 08:29, 29 May 2007 (UTC)
- duplication is badbadbad. --MarSch 10:53, 29 May 2007 (UTC)
- ...yes but
(WP:PAPER )until that December date there was no real "Linux OS" per se as it still needed MINIX. The release of the kernel is an important date but more important is when it became self-sufficient from what it bootstrapped itself from. Ttiotsw 11:01, 29 May 2007 (UTC)
- ...yes but
Linux is not an operating system
Linux is not an operating system. It is a kernel. Why does it say it is an operating system? --212.247.27.91 14:10, 29 May 2007 (UTC)
- Do us a favor and read the archives. -- AdrianTM 14:36, 29 May 2007 (UTC)
- What date? --212.247.27.197 13:17, 1 June 2007 (UTC)
- Any. -- AdrianTM 13:37, 1 June 2007 (UTC)
- All of the archives contain this discussion. I think, actually, that most of the discussion has been about this, and the GNU thing.-Localzuk(talk) 14:17, 1 June 2007 (UTC)
- Any. -- AdrianTM 13:37, 1 June 2007 (UTC)
Maybe we should document the reasons to make it easier. Mack the Turtle 04:23, 3 June 2007 (UTC)
- Huh? They are documented - in the archives.-Localzuk(talk) 19:29, 13 June 2007 (UTC)
Games
The article states that there are far less games for Linux than for Mac OS X. Is that correct? (IANA proffesional, but this looks suspicious.) 193.171.249.108 09:48, 6 June 2007 (UTC)
- I suppose it depends on what it is using as its context. Is it meaning 'modern, commercially released games' then it is likely true. But if it includes hobby games and freeware I would say not. It needs a source either way.-Localzuk(talk) 09:53, 6 June 2007 (UTC)
Date for just one component: glibc!
The intro currently says the Linux kernel was released on whatever date. I added a date to the sentence about GNU, but the date was deleted "to keep the intro concise".
This is an article about the operating system. There is no reason to give a release date for version 0.1 of the kernel but not for any other component of the system.
What's the solution? Remove the kernel release date? Or, as a temporary solution, is it ok to re-add the date to the GNU project sentence? Gronky 09:50, 13 June 2007 (UTC)
- There was no project called "Linux" prior to Linux 0.1. This is why the kernel has a date in the intro. Please don't drag us round another dance about "assigning due weight" to various random GNU components. It does not benefit the article in the slightest. Chris Cunningham 10:06, 13 June 2007 (UTC)
- This is not the article about that project. This is the article about an operating system. Release dates for individual components are off-topic and misleading as to the topic of the article. Gronky 10:39, 13 June 2007 (UTC)
- As pointed out to you a dozen times, this is not an article about GNU/Linux-but-just-called-Linux. The whole GNU/Linux thing is a construction of the FSF. This article is about the operating system called Linux. So far as most of the world (including most computer professionals, and very possibly a majority of the free software community) is concerned, there's no clear distinction between Linux-the-OS and Linux-the-kernel, so what was released with Linux 0.1 was an operating system. Chris Cunningham 11:07, 13 June 2007 (UTC)
- I didn't even mention "GNU/Linux". Whether or not most of the World differentiates between the OS and the kernel, they are not the same, and Wikipedia does differentiate (even if many pages use confusingly identical names for the two). The release date for the kernel belongs on the kernel page, not here. Gronky 12:27, 13 June 2007 (UTC)
- I'd argue that they are the same insofar as establishing point of origin goes. Hence the date. Chris Cunningham 14:28, 13 June 2007 (UTC)
- Huh? GNU has a point of origin, the Linux kernel has a point of origin which is later, and the OS made from the combination has another point of origin which is later again. Linux kernel was not made for GNU, and GNU was not made for Linux kernel. They cannot share a point of origin. Gronky 14:39, 13 June 2007 (UTC)
- Yawn. This is not an article about GNU/Linux-but-just-called-Linux. The distinction between OS and kernel is an artificial construction of the FSF. The only reason this isn't pushed is because there aren't many people wandering Wikipedia demanding that "Minix/Linux" be given equal weight in articles because Linux used to depend on the Minix userland. Chris Cunningham 15:38, 13 June 2007 (UTC)
- "The distinction between OS and kernel is an artificial construction"? What's the thing I download at kernel.org? An OS? What's the thing I get from gNewSense or Debian? Just a kernel? There is obviously a distinction. Gronky 17:34, 13 June 2007 (UTC)
- The thing you download at kernel.org is the Linux kernel, the things you get from Debian etc... are distributions of the Linux OS.-Localzuk(talk) 17:52, 13 June 2007 (UTC)
- Thank you. One is a kernel, one is an OS. Gronky 17:58, 13 June 2007 (UTC)
- The thing you download at kernel.org is the Linux kernel, the things you get from Debian etc... are distributions of the Linux OS.-Localzuk(talk) 17:52, 13 June 2007 (UTC)
- "The distinction between OS and kernel is an artificial construction"? What's the thing I download at kernel.org? An OS? What's the thing I get from gNewSense or Debian? Just a kernel? There is obviously a distinction. Gronky 17:34, 13 June 2007 (UTC)
- I'm not entering back into this. Try editing the Inkscape page to start with "The GNU Project, with the goal of creating a Unix-like, POSIX-compatible operating system composed entirely of free software,[8] had begun development in 1984, and a year later Richard Stallman had created the Free Software Foundation and wrote the first draft of the GNU General Public License (GPLv1)" because Inkscape uses glibc. The primary distinguishing component of the OS, the bit that gives it its name, was first released in 1991. Not mentioning this in the intro seemed strange. Mentioning other random dates for the overt purpose of confusing people is not good practice. Chris Cunningham 18:37, 13 June 2007 (UTC)
- I've said your there's no distiction between OS and kernel position is unsupportable, someone agreed, and now you're saying you won't explain yourself any further. Fine. Gronky 19:20, 13 June 2007 (UTC)
- The Inkscape page is about a program that could be called part of an OS (the "Linux" OS), just as the Linux kernel is, no matter how relevant the kernel is or is not for the OS (remember this old adage of not necessarily giving due weight. Yes, the one argument you use against the "GNU/Linux" name). So yes, the page for Inkscape should mention its creation date, as well as Linux kernel's should mention its creation date. Now the page for the "Linux OS" should mention the creation date for the "Linux OS", not the one of the kernel, of Inkscape or of udev. It is so annoying to have people repeating that this page is about an OS (because allegedly the OS's name is "Linux", not "GNU/Linux"), then forgetting that the page talks about the freakin' OS, not the kernel. I thought that using "Linux" for the name of the OS was just utilitarian, pragmatic, not hiding an agenda of mistaking what an OS and what a kernel is, then pretending the kernel is the whole OS, then pretending the OS was created in 1991, then pretending that Linus Torvalds created the OS. Please, prove me that I am wrong in fearing the latter. — isilanes (talk|contribs) 14:04, 14 June 2007 (UTC)
- As far as tracing the origins of 'Linux' the OS, the release date of the Kernel is the starting point - without it, there wouldn't be an OS called 'Linux'.-Localzuk(talk) 19:28, 13 June 2007 (UTC)
- The argument "without it, there wouldn't be an OS called 'Linux'" can be used to support adding dates and more information about numerous important components. Gronky 11:43, 14 June 2007 (UTC)
- Unsuccessfully it seems. The offer to go and change Inkscape's release date to 1984 stands. It'd be more entertaining than continuing to waste electrons pushing the same old argument on Talk:Linux. Chris Cunningham 12:01, 14 June 2007 (UTC)
- Whatever Inkscape change you're for or against, discuss it there, not here. Gronky 13:05, 14 June 2007 (UTC)
- Chris, saying that the release date of Inkscape is that of a component of it (glibc), regardless of it not matching the date when a group of people joined efforts to release the program itself (Inkscape) is just about as idiotic as saying that the release date of the Linux OS is that of a component (the kernel), regardless of it not matching the date when a group of people (the GNU movement) joined efforts to release the OS itself (Linux, the free software Unix-like operating system, which they called GNU... but the name is irrelevant). Thanks for giving support to our arguments. — isilanes (talk|contribs) 14:04, 14 June 2007 (UTC)
- I say remove the kernel release date. If anything, the intro should mention when the first distributions started appearing. 80.233.255.7 22:46, 13 June 2007 (UTC)
- I say keep the date. A real publication date was an important event in the world's history (although we didn't know it at the time). --Alvestrand 23:34, 13 June 2007 (UTC)
- Same can be said about the X Window System or, for that matter, any major piece of software that you see on a GNU+Linux system. In fact, your argument supports mention of the GNU C Library's release date just as well. But that would be missing the point. This isn't an article about the individual components, it's about the combination of those components.
- Note that I'm talking about the intro. 80.233.255.7 08:57, 14 June 2007 (UTC)
- Exactly ("This isn't an article about the individual components, it's about the combination of those components.") Gronky 11:45, 14 June 2007 (UTC)
- But the point is that the release of the Linux kernel was the 'first step' to this OS. It wouldn't be 'Linux' without that kernel - it would be something else, but that something else may well still have used glibc, x-windows, gnu-tools etc... As a starting point, the date is important when describing the history. The first 'distro' should also be dated too-Localzuk(talk) 12:22, 14 June 2007 (UTC)
- The kernel wasn't developed to make this specific OS happen. If you call the development of projects which weren't developed to make this OS happen a step toward making it happen, then kernel was not the "first". Gronky 13:05, 14 June 2007 (UTC)
- The statement that "The kernel wasn't developed to make this specific OS happen" is a flat-out lie. I expect you to retract it if you're serious arguing in good faith and not just trolling. Chris Cunningham 14:01, 14 June 2007 (UTC)
- Huh? That's undisputed fact. Linus wrote one core component, for fun. He never set out to make the OS that this article is about. Gronky 14:09, 14 June 2007 (UTC)
- Gronky, it does not matter about intent. It doesn't matter if Linus set out to start the OS. What matters is that the Kernel was produced and that is the historical start point for this OS. There would not be an OS called Linux without that kernel being released - it would be a different OS. That is the single start point of this OS.-Localzuk(talk) 16:13, 14 June 2007 (UTC)
- Huh? That's undisputed fact. Linus wrote one core component, for fun. He never set out to make the OS that this article is about. Gronky 14:09, 14 June 2007 (UTC)
- "There would not be an OS called Linux without that kernel being released". Why not? I do not get it. I though we had long ago agreed that the name "Linux", in itself, was irrelevant, and that the OS could as well be called "Bananas", because the name was a matter of use and pragmatism, not recognition to major components of it. Everyone calls the OS "Linux"? OK, then for all purposes in Wikipedia, the OS will be referred to as "Linux". Hence the name of this article. I didn't know that by choosing a given name for the OS we were shifting the very subject of the article: the free software Unix-like operating system developed by the GNU movement (incidentally using a kernel developed by Linus Torvalds), which eventually became what is known as "Linux". Unless, of course, you are denying that this Linux is that Linux. — isilanes (talk|contribs) 20:59, 14 June 2007 (UTC)
- Go write a history section on GNU C Library. It needs one. --Alvestrand 21:27, 14 June 2007 (UTC)
- Right, the OS called 'Linux' is this OS. It is an OS that was developed around the Linux kernel produced by Linus Torvalds. It was not created by GNU and it was not created by any other single project - it was created by a variety of projects (including X-Windows and TeX). It has always been called Linux, and likely always will be called Linux - as it was built around that particular kernel. Now, if it had been developed around, lets say, 'Hurd' then this OS would not be 'Linux' it would likely be 'Hurd' and the initial date would be the first release of that kernel. The single most important part of any OS is the kernel - this is undeniable from any POV, be it a computer science background or a historical background.
- Let me put it this way, if Linux, the kernel, had not been released, what would GNU have used to create their OS? It wouldn't be the same OS had they chosen any other kernel.
- Arguing over semantics is getting very old - nearing a state of stupidity. Hundreds of hours have been wasted arguing over the whole 'Linux'/'GNU/Linux' nonsense and it is obvious which term is more commonly used and it is obvious that the only reason to include it is POV pushing by the GNU camp - that is unless you can provide compelling evidence to support the claim that GNU created Linux.-Localzuk(talk) 22:36, 14 June 2007 (UTC)
- Had it been developed around Hurd, it would likely be called "GNU", not "Hurd". And it would be called so because Hurd would have been the last element to make the GNU OS (which was being developed from 1983 on) complete. Now, for some obscure reason, since the Linux kernel, and not the Hurd, was that last element, it seems that not only the same project (the GNU project) should be named differently, but even the nature of the project should be understood differently: now the OS is regarded as being built around the kernel, instead of the kernel being considered the last (and most important, maybe) stone in the building. We are not discussing semantics here. — isilanes (talk|contribs) 15:23, 15 June 2007 (UTC)
- Nobody is arguing about "it should be called GNU/Linux" in this thread. That's a separate issue and the topic of discussion is complex enough without adding it. The OS was not developed around the kernel. Most of the OS was developed before Linus's project even began. The other OSes the GNU system would have been used to create if not for Linus's kernel is a speculative topic (for which I have answers, but I'm trying to keep this on-topic) that isn't important for this article. Gronky 11:25, 15 June 2007 (UTC)
- By any metric whatsoever, the majority of the code in a modern Linux distro post-dates the release of the kernel and was written primarily for use with the "Linux OS" as defined by the kernel. It is sheer folly to suggest that a compiler, C library and a set of cloned userland programs (the extent of GNU pre-Linux) represents a majority of what makes up a non-toy Unix. The only people who would do so are zealots and people stuck in 1983. Chris Cunningham 12:27, 15 June 2007 (UTC)
- By any metric, modern Linux kernels have little to do with the works of Torvals (I think it was something like 2% LOC are Linus's). So by that logic, what we know call "Linux" is not that kernel of 1991. Don't play the "major contribution" card. It might backfire. By any metric, the creation of the GNU movement was infinitely much more of a breakthrough than the release of a usable kernel by Torvalds. The founding stones of the OS (the Linux OS) were set in 1983. The philosophy behind free software (unless you want to dissociate Linux and the free software. The free software under whose GNU GPL license the Linux kernel is) was developed in the 80s. The roadmap to what the OS (the Linux OS) should look like was sketched in the 80s. Gosh, in 1991 the Linux kernel was recognized as the "final piece" that would make the OS envisaged by the GNU movement work because there already was everything (essential) else. It is so embarrassing to read that "the Linux OS was built around the kernel", when it is undisputed that what the kernel did (unplanned for, and unnoticed at first for its maker) was to fill a niche in a greater scheme of things (the GNU movement). — isilanes (talk|contribs) 15:23, 15 June 2007 (UTC)
- It is not "undisputed" at all. That version of events is a fabrication of the FSF; the article currently gives the generally-accepted version, which was that while GNU meant Torvalds had a free editor and compiler to work with, the kernel is and always has been the centre of the entire OS project. And don't be so obviously dishonest as regards the name; it doesn't matter how many LOCs Linus himself has contributed to the Linux kernel, because unlike Richard Stallman the vast majority of free software contributors don't think projects should be renamed based on who's hacking on them. The kernel as a whole is a substantial chunk of the LOCs in a generic Linux distro, especially when you then add on those parts of the OS which were specifically designed to work with said kernel in the first place (such as HAL and ALSA). Chris Cunningham 15:28, 16 June 2007 (UTC)
- People wrote software for the OS, but the OS was not defined by the kernel. To application programmers, the OS is more defined by glibc. So now you're saying glibc is the main component. Gronky 12:55, 15 June 2007 (UTC)
- I'd love to point out here that glibc is a prime example of anti-GNUism, given that Ulrich Drepper is much harder on GNU bullying than Linus is. But hey, Talk:Linux will continue to be a wasteland of GNU/pushing until the end of time. The important thing is to ensure than casual editors are aware of the minority status of such arguments. Chris Cunningham 01:07, 15 June 2007 (UTC)
- Which glibc do you mean, Chris? The GNU C Library? The one under the GNU General Public License? — isilanes (talk|contribs) 15:23, 15 June 2007 (UTC)
- Yes. Go do some research. Chris Cunningham 15:28, 16 June 2007 (UTC)
- Sorry, the link you give must be wrong: I only see the rants of a person who opposes Stallman's attitude. Drepper is not glibc, and Stallman is not the GNU movement. — isilanes (talk|contribs) 19:03, 17 June 2007 (UTC)
- The realist part of the point was that it's ridiculous to try to add the release date of glibc to the LINUX article when it's not in the GLIBC article. Adding it there is uncontroversial. People who want to add it to the Linux article but refuse to take the time to add it to the GLIBC article are showing that they don't care about making information available to the world, they care about making a point. Let's build an encyclopedia. --Alvestrand 17:57, 17 June 2007 (UTC)
- Drepper is glibc. He's the one who ported it to Linux. He's the lead developer and maintainer. As a primary source, his opinion is several fathoms more important than that of random wikipedians. It shall continue to be afforded that precendence in this article. Chris Cunningham 19:18, 17 June 2007 (UTC)
- Funny that the glibc article itself (it might be wrong) says: "Glibc was initially written mostly by Roland McGrath, working for FSF in the 1980s". Drepper is just the current maintainer and lead contributor... he didn't "invent" the stuff, and he doesn't "own" it. He ported it to Linux? Fine. He made it work with a given kernel, which is not the same as "creating" the software, which arguably existed before, right? We all agree that glibc was created from scratch by the FSF, right? We all agree that it is one more GNU piece in the GNU scheme of how to make a GNU OS, right? Maybe that "GNU scheme" didn't succeed. Maybe we should pray to Torvalds before going to bed, and praise him because Linux (the OS) wouldn't exist without him. Believe me, I have no problem with that. Moreover, I can and will praise Drepper hereafter for his contributions to free software. However, we are not here to discuss Drepper's or Stallman's contributions, or whose d*ck is bigger. For all accounts, one or both could be total a**holes, discussing like stupid kids (as your link confirms). So what? The only relevant point is that the OS we now know as Linux is the matured result of a project started in 1983, not 1991. We can give dates for when it was functional, or when it was complete, or when it was mature or when it became modern. Maybe any or all of these steps are more important for Linux and FS than the conception of GNU's Not Unix by the GNU movement. It is true that for all these steps, people like Torvalds and Drepper have been crucial, and their contributions don't go unnoticed. But the project was not started by them, and neither by people before the FSF. It was the FSF and the GNU movement in the early 80s, lead by Stallman. Don't let the name "Linux" fool you! — isilanes (talk|contribs) 08:55, 18 June 2007 (UTC)
- You're making this up as you go along, and I'd appreciate it if you remained civil (or, if you're going to use words like "dick" and "asshole", didn't pretend that they aren't offensive if they have wildcard substitution). Drepper is not "just" the current maintainer, he's the guy who ported glibc to Linux in the first place - against the wishes of the FSF. Linux is not "a variant of the GNU system", it's a competing OS which happens to contain some of the same parts, and the FSF's recent attempts to co-opt the project for the benefit of their egos bear a reasonable similarity to the trick Ray Kroc pulled. The Linux operating system originated with Linus Torvalds, and the prior existence of GNU is pertinent only in the same indirect way as the prior existence of C. Chris Cunningham 09:30, 18 June 2007 (UTC)