Talk:Linux/Archive 19

This is an archive of past discussions about Linux. 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 17

→

Embedded Devices: Yamaha Motif

Is there any reference to prove that the Motif really runs Linux? ~~I thought the claim was interesting, but I cannot easily find proof of it online.~~ I've found a PDF (http://www.mvista.com/downloads/Case_study_MontaVista_Linux_and_Yamaha.pdf) which shows that the Motif XS line of products uses MontaVista Linux. I will edit the article accordingly. Greenlead 16:21, 13 August 2007 (UTC)

Arbitrate or remove kernel date

I, for a second time, replaced the release date for one component (the kernel) with an approximate release date for the OS. This has been reverted by User:Thumperward with the edit summary "don't make me bring this to arbitration. this hasn't been accepted in talk". Ok, if you really think you have a case, then arbitrate. Gronky 14:55, 17 June 2007 (UTC)

I hate having my bluff called. Then again, I suppose if it's going to stop this talk page being spammed every three months by the same editors making the same arguments, a ruling on this would be nice. I'll get round to it. Chris Cunningham 15:09, 17 June 2007 (UTC)

Well Gronky, you don't have consensus for the change so your unilateral edit is not acceptable. I think the first stage is an RFC - but what on? Gronky's behaviour or the content itself?-Localzuk^(talk) 15:26, 17 June 2007 (UTC)

A little of both. I really wish it didn't have to come to this; Gronky is one of the most active free software contributors on here and we've collaborated on improving a good few articles, and outside of the Linux article itself we haven't warred about this to my knowledge. But we can't really have a situation where hundreds of kilobytes of discussion are generated on rebutting the same few arguments every three months or so, not least of which when it results in edit wars. There's seemingly nothing which can be said to resolve this situation unofficially (and lord knows we've been over it for long enough). Chris Cunningham 15:31, 17 June 2007 (UTC)

I won't take the outcome personally. I recommend wording the arbitration (or RFC) request as tightly as possible so that (a) we get an answer to the actual question asked, rather than about some other disagreement such as "GNU/Linux" vs. "Linux", and (b) to give the arbitrators no more work than we need them to do - they are likely to have less background knowledge and less time for this discussion than we do.

FWIW, a definition of the disagreement off the top of my head is: Should the kernel be mentioned (a) more prominently and (b) in more detail in the intro than the software of the GNU project.

If you add pro/against arguments, then, I think my points, in somewhat order of importance, are that the GNU contribution is the larger of the two, it had eight years of work put into it before the Linux kernel even began, it is the one of the two that was done with the aim of making an OS, the GNU libc API is what application developers write code for, GNU was the first work that was done to make such an OS exist and was thus the "first step" in making the OS, and GNU enabled the Linux kernel development by providing a compiler and a licence. Gronky 17:30, 17 June 2007 (UTC)

The whole point is that we are not the ones to make the decision. We have to report on only that which has previously been collected, and assign due weight to relative arguments where there is no accepted normative answer. There is no accepted arbiter to declare either said "correct" in this case. There is ample historical documentation to lend weight to the argument that GNU as a project had stalled prior to Linux and that the Linux kernel served as a locus of development almost from the moment it was released from sources which are affiliated with neither Torvalds nor the FSF. There is little historical documentation to counter this argument which comes from sources not affiliated with the FSF. Chris Cunningham 19:01, 17 June 2007 (UTC)

I find this discussion a bit funny. The cusp here is neither when was the kernel operational - nor whether GNU is an important part. Its rather simple: When was the first Linux operating system, independant of Minix released. The kernel has its own history (and article) - record the release of 0.1 there, GNU has its own history, record the release of the main components there. Now the first Linux OS (iirc) was released from funet.fi, by one of Linus' friends (can't remember the name here - but he was the one who talked Linus out of the Freeks name (hope i'm remembering this correct)). That the OS was released with GNU can be considered (depending on viewpoint) a lucky accident (linux viewpoint), or a part of the picture (gnu viewpoint). But in this particular context, it should be the first instance of a working independent linux system (ie. kernel + basic userspace tools (init,shell etc)). If this is GNU, BSD or Minix based is rather irrelevant. Just my 2 cents. --Kim D. Petersen 19:56, 17 June 2007 (UTC)

Thing is, this is extremely subjective. Obviously some people could boot into something that they called "Linux" at basically any point after the initial release; we're getting into very murky and unsourced territory here. As has been extensively argued, it is impossible to nail some point where the "Linux OS" (as opposed to "the kernel") achieved existence. It was a gradual process. The only definitive start point is the release of the kernel, prior to which is was absolutely impossible to boot into something called "Linux" (and indeed, impossible by boot into something called "GNU"). Chris Cunningham 00:23, 18 June 2007 (UTC)

No date can be given, and it came into existence gradually - we agree there. Why not put this in the intro? It could be attributed to a year, or a time such as "around 1992". Gronky 00:36, 18 June 2007 (UTC)

Still waiting on this. Gronky 11:42, 20 June 2007 (UTC)

Sorry, I was ignoring you, which seemed to be the appropriate response to your 180-degree misrepresentation of my position. The current wording is fine. Muddying the waters serves no purpose other than to confuse casual readers. Chris Cunningham 11:51, 20 June 2007 (UTC)

You've misunderstood something somewhere. I'll say it in an umabguous way. You said: "it is impossible to nail some point where the "Linux OS" (as opposed to "the kernel") achieved existence. It was a gradual process." So I'm suggesting adding that to the intro - word for work if you like (although the parenthesised part could be removed). Gronky 12:00, 20 June 2007 (UTC)

Heh. Now we're getting existential :) Allow me to present an analogy. It is still a point of contention as to when life begins. However, there is reasonable consensus that you don't start counting how old someone is until they're born. Linux as an operating system is an organic bit of code where a distinct point of formation is hard to put down, but if you're going to provide a date for someone to track it with, the original release of the kernel is a very common one. Chris Cunningham 12:13, 20 June 2007 (UTC)

Life, inkscape, etc. aside. Do you have any references saying how common it is to give Sep 17th as the "birth" date of the OS? Gronky 12:23, 20 June 2007 (UTC)

Again, there's nuance, but Lars's history (which, by the way, is an excellent resource we should be using) states "Into this scene came Linux, in October, 1991" where "Linux" is being used in the OS sense. I'm tempted to write "October" off as poetic licence. Torvalds obviously uses "kernel" and "OS" interchangeably in his own histories. Googling for "linux birthday" obviously supplies 17 September 1991 pretty universally, though separating that into "kernel" as "OS" is an exercise left to the reader. The point is that if you have to pick a date for the sake of keeping things simple, there's only one date in it. Chris Cunningham 12:59, 20 June 2007 (UTC)

Lars is clearly talking about the kernel: "...Linux, in October, 1991. Linus Torvalds, the author...". About "if you have to pick a date", you don't. As you say, it was an evolution. We could surely agree on a year, but a specific day, no, and there's no need. Gronky 13:31, 20 June 2007 (UTC)

By reading Linus' mails, its quite clear that the system wasn't up to OS status before 0.11 - or possibly 0.12 (the first that came with more features than the "essentials"). That would place the "birth" of the OS between 19 Dec ~~2001~~1991 and 5 May ~~2002~~1992. --Kim D. Petersen 14:11, 20 June 2007 (UTC)

I assume you mean 1991/1992, Kim :)

Gronky, if removing the words 17 September will kill this silly thread for another couple of months I'm wholeheartedly in support. Chris Cunningham 14:57, 20 June 2007 (UTC)

~~Sure. It's not what either of us wants, but at this level of talk-tiredness, it's enough to make me give up this thread. Gronky 10:11, 21 June 2007 (UTC)~~ How could that be done? (I just realised that the sentence would be no less kernel-centric and would be silly if the date was just deleted) Gronky 10:42, 21 June 2007 (UTC)

(replying to further up comment at this indent level) This "GNU stalled pre-Linux" claim is new to me. Accounts such as Michael Tiemann's[1] and the GNU-published but independently-verifiable GNU's bulletins, from 1986 to 1998,[2] refute it and I don't know what backs it up. But to get back to the point, the current selection and presentation of information in the intro is overly (and needlessly) kernel-centric and downplays other contributions, in particular the very large GNU contribution. Gronky 00:36, 18 June 2007 (UTC)

I could have sworn that you specifically said this wasn't about GNU earlier in this thread. This is yet another quagmire, and you've failed to convince anyone to change their minds. Chris Cunningham 11:51, 20 June 2007 (UTC)

You're mis-remembering. I said this isn't about "GNU/Linux" vs. "Linux". Naming is a seperate thing. Whether it's name "Linux", "GNU/Linux", or "bananas", it should still accurately reflect the contributions that made it happen and it's history. Gronky 12:23, 20 June 2007 (UTC)

Reset Indent

Anticipating this argument a few weeks ago I had proposed that we mention...

"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."

it didn't really get traction due to something about duplication. IMHO it doesn't matter when the kernel was released but from an OS POV it was when it was a stand-alone OS (with GNU stuff etc).

That December date with version 0.11 is really the only date that anyone can say any "Linux OS" came into existence. It was from that date forward that the plethora of distros and ports to alternate archs were possible. Ttiotsw 20:40, 17 June 2007 (UTC)

That doesn't resolve the question at hand, though. Chris Cunningham 00:23, 18 June 2007 (UTC)

Naming

In the article, everything under the section titled "Naming" is about pronounciation, therefore, shouldn't it be renamed something like "Pronunciation"? Sorry if this has come up already. Penman 1701 02:36, 18 June 2007 (UTC)

i have another thought about naming Linux a operating system. Isn't it more of a platform then OS? Because Linux is actually kernel, and kernel only is not really a OS...well maybe it is, but still wouldn't platform be a better description to Linux? psycho_NIX 07:12, 14 August 2007 (UTC)

What's a "platform"? Gronky 11:05, 14 August 2007 (UTC)

hmmm. Well this is what my dictionary says about it:

Main Entry: plat·form

Pronunciation: 'plat-"form

Function: noun

Usage: often attributive

Etymology: Middle French plate-forme diagram, map, literally, flat form

1 PLAN, DESIGN

2 a declaration of the principles on which a group of persons stands;

especially : a declaration of principles and policies adopted by a

political party or a candidate

3 a (1) : a usually raised horizontal flat surface; especially : a

raised flooring (2) : a device or structure incorporating or providing

a platform; especially : such a structure on legs used for offshore

drilling (as for oil) b : a place or opportunity for public discussion

4 a a usually thick layer (as of cork) between the inner sole and

outer sole of a shoe b : a shoe having such a sole

5 a a vehicle (as a satellite or aircraft) used for a particular

purpose or to carry a usually specified kind of equipment b : OPERATING

SYSTEM;

Simply said something (specific) you use to base other things upon.

Would it be wise to include an analogy to explain that a 'Linux system' or the 'Linux operating system' is like an atom with the nucleus being the Linux program and other programs (whether GNU or not) function in orbit to this? The GNU/Linux label could then be said to refer to the whole atom, and 'Linux' to the nucleus. Ken Lewis81 14:02, 12 September 2007 (UTC)

Infobox Mascot

How come the picture of Tux in the infobox isn't his normal rendition? Instead it's his Crystal version... Am I the only one that thinks that should be changed? Cosmotron 20:29, 25 June 2007 (UTC)

With "normal", you mean "traditional"? I liked the Crystal version, and see no reason why the logo can't evolve (unless it is somehow an "official" logo, and we have to stick to an exact version). It was changed back, it seems. I don't personally care too much about it, anyway. — isilanes (talk|contribs) 09:31, 27 June 2007 (UTC)

Yes, by "normal" I meant the "traditional" Tux that is now being displayed. -- cosmotron 22:21, 3 July 2007 (UTC)

As this article widely acknowledges that Tux is the mascot of the kernel, and that 'Linux' used here refers to what some might called GNU/Linux, wouldn't an image that reflects this, such as Gnu-and-penguin-color.png, be more appropriate? --Mattl 01:04, 31 July 2007 (UTC)

No. -- AdrianTM 01:35, 31 July 2007 (UTC)

No... Considering that Tux is used to represent the OS also (for example, it features on the Unreal Tournament 2004 box to indicate Linux support).-Localzuk^(talk) 07:52, 31 July 2007 (UTC)

Isn't that just adding to the confusion though? To quote the article the logo links to 'Tux (also known as Tux the Penguin) is the official mascot of the Linux kernel.' --Mattl 00:09, 1 August 2007 (UTC)

That's not our fault, so to speak. "Linux" as an entity is defined entirely by how others treat it. There's nobody to name an official mascot for "Linux", but practically everyone accepts that un-officially, Tux represents the OS as well.

Tux is really the closest thing Linux has to a logo. People see him, and think "Linux". Anyone that's familiar enough with linux to even begin to think of a logo will think of Tux. Tux shows up everywhere- on software boxes, on boot screens, on CD labels. There's really no reason to change it --Longing.... 13:44, 1 August 2007 (UTC)

Hardware compatibility

I'm surprised I can't find any info on hardware compatibility. Wikipedia can't list all the hardware and their compatibility, but there should at least be an external link to a site that does (which is what I was looking for). And there should be some general story about the sorts of hardware that is supported better or worse and the reasons. Such as that harddisks are always (?) supported, but that drivers for less general and more recent hardware are often not completely functional, if they exist at all, because the Linux community has to write it, often without having the relevant technical specifications, whereas msWindows-drivers for the hardware are written by the hardware manufacturers themselves. I won't write this myself because I don't know enough about it. But there should at least be an external link to a site with all the details. I assume there is such a site.
Or am I looking in the wrong place? If so, something should be done about that, because I'm an experienced Wikipedia user, and if even I can't find it .... Not that I've looked too hard, but then I shouldn't have to. I simply searched for the word 'hardware'. DirkvdM 07:42, 27 June 2007 (UTC)

It used to be here but people don't want to make Linux look bad so it's gone.

Deemphasize history?

I just caught this ZDNet blog post, which brought up the point that the things that people might look for from this article aren't currently emphasized. In particular, he doesn't seem to like the fact that the history is put above any actual description of what Linux actually is (other than the short literal definition in the lede). While we certainly don't want to "get more Windows users to switch to Linux", of course, the point may be something to consider. —AySz88 \^-^ 07:24, 2 July 2007 (UTC)

There are a few problems with this. The first is that it's only possible to talk about what an operating system is for so long before the article reads like a condescending advert. As you say, we're really actively trying to avoid targeting this at curious Windows switchers. The second is that Linux really is different things to different people, and that the blog post in question is really targeted pretty directly at "Ubuntu" of "GNOME" or whatever rather than "Linux". I think the best solution to this is to clean up the wording of desktop distros in the intro. Chris Cunningham 11:55, 2 July 2007 (UTC)

"Criticisms" section

I, per Wikipedia:Criticism, which states the following:

In general, making separate sections with the title "Criticism" is discouraged.

propose to remove the section entitled "Criticisms" and merge the content that presently exists within to sections that are more appropriate and are directly related to the material. I believe this would improve the quality of this article. Please give your opinion, if you care to provide it, so that we may reach consensus on what we should do regarding this issue. Iccdel 02:15, 6 July 2007 (UTC)

Hear, hear. ¦ Reisio 04:10, 6 July 2007 (UTC)

Totally agreed, what's next "Praises" section? --AdrianTM 05:24, 6 July 2007 (UTC)

Waves hands in the air in typical silent consensus reaching manner* (aka. Agree!) -Localzuk^(talk) 12:19, 6 July 2007 (UTC)

That makes criticisms a lot harder to find though... :S

~~Lazyguythewerewolf . Rawr. 17:25, 13 September 2007 (UTC)

Unix-Like ?

Being familiar with *nix's I know how linux fits in: otherwise, I would be misled by 'unix-like' which would imply immitation, but not necesarily compatibility.

The introduction would benefit from a refinement of 'unix-like' to reflect less ambiguously what 'like' means: for example, posix, portability due to common methods and a common infrastructure and so forth (e.g. redirction, piping - including the exact syntax as well as the implementation). 66.245.28.149 21:17, 7 July 2007 (UTC)

"Unix clone" would be better? Actually it's not a clone since it differers in some aspects (even improving over Unix) -- AdrianTM 22:18, 7 July 2007 (UTC)

The term "Unix-like" is pretty common. It's reasonable to expand this as meaning "largely POSIX-compliant" or "Unix-compatible", as long as it's clear what's meant by "compatible" ... --FOo 03:52, 8 July 2007 (UTC)

Yup. Unix-like goes over this in some details. The GNU folk have been using "Unix-like" in this sense for over twenty years now, so it's not like it's a neologism. Chris Cunningham 09:45, 8 July 2007 (UTC)

agree @ Chris Edgesurge 03:57, 27 August 2007 (UTC)

Successful good article nomination

I am glad to say that this article which was nominated for good article status has succeeded. This is how the article, as of July 12, 2007, compares against the six good article criteria:

1. Well written?: It is well written

2. Factually accurate?: All sources are stated

3. Broad in coverage?: Provides Coverage on major aspects.

4. Neutral point of view?: I did not see any bias, however the criticism needs to be bigger

5. Article stability? Lot of traffic on this article, however, it does not change constantly.

6. Images?: Images helps the article, "A Picture is worth a thousand words"

If you feel that this review is in error, feel free to take it to a GA review. Thank you to all of the editors who worked hard to bring it to this status. — -Flubeca^Talk 01:06, 12 July 2007 (UTC)

I'll just echo Wikipedia:Featured_article_candidates/Linux/archive1 - David Gerard's comment. History seems to repeat itself constantly with this page, unfortunately. RN 03:17, 12 July 2007 (UTC)

Just for fun?

excuse my ignorance (I'm a 'noob' at this), should there be some reference to Mr. Torvalds writing this "just for fun"(apparently his own words)? -it seems relevant as this was what got me interested me in the subject in the first place. —Preceding unsigned comment added by The evil jelly (talk • contribs) 01:21, 14 July 2007

it is. See the wikiquotes link at the bottom. -_L^augh! 10:21, 14 July 2007 (UTC)

He also needed the linux to support the cheaper i386 arcitechture, by many known as the first home computer that where easy to get hold of.

History section

Firstly, this talk of WP:CON is highly amusing. I've been involved in pages and pages of discussion here and the current history section got the way it did after a very large amount of discussion. Bold, contentious reworking of the history section to convey a different impression of the subject is unwarranted at this stage. To revert to such a version while making claims of consensus makes no sense whatsoever. Chris Cunningham 13:52, 25 July 2007 (UTC)

There's no need to add more headers. The section is fairly short, this is a summary article, and it reads fine as-is.Chris Cunningham 13:52, 25 July 2007 (UTC)

Yes, there is definitely a need for more headers. They separate what came before linux (which is vital in order to understand where Linux was coming from), and where Linux itself began. --L^augh! 14:19, 25 July 2007 (UTC)

There's no need to do this. It's a summary article and the history consists of only four short paragraphs.Chris Cunningham 14:48, 25 July 2007 (UTC)

The GNU logo only serves to advance the whole "Linux is GNU" argument, which isn't the consensus conclusion.Chris Cunningham 13:52, 25 July 2007 (UTC)

I agree, and not to mention it's way too large anyway. I was planning on removing it. This is a case where trying to improve on someone else's edits will get you farther than edit warring. I might put work-in-progress on the page for a bit while I work on that section. I DO think there are some things done better in your version, and I'd rather take the best parts of both of them than just act like one is better than the other and that's that. --L^augh! 14:19, 25 July 2007 (UTC)

That's nice, but removing a disagreeable edit which wasn't there this morning isn't "working towards consensus".Chris Cunningham 14:48, 25 July 2007 (UTC)

I think they (most of it's from the German Wikipedia) just put it in there because they were talking about GNU. Mike92591 22:31, 26 July 2007 (UTC)

Most importantly is this "filled a gap" nonsense. GNU was not some finished operating system waiting for someone to plug in a kernel. It was a collection of Unix software which by 1991 needed a direction. The driving force was the Linux kernel, and the existence of code which helped spur on Linux was an added benefit.Chris Cunningham 13:52, 25 July 2007 (UTC)

I don't see what you're talking about, but again, this could be easily solved by working WITH other editors instead of acting like you are better than them. Now please, stop edit warring, and start edit conf'ing. Reverting to the same version over and over again never got a Wiki anywhere --L^augh! 14:19, 25 July 2007 (UTC)

See the top of the current talk page, or the end of the last one. Chris Cunningham 14:48, 25 July 2007 (UTC)

GNU was some operating system waiting for someone to plug in a kernel. At the time there really was no other free and useful kernel (hurd sucked,BSD was not legitimate). Linux certainly filled the gap for a free OS. Mike92591 22:30, 26 July 2007 (UTC)

UI

The lack of a subheader does not confer any sort of default acceptance. I have no idea where you're getting this. Chris Cunningham 14:10, 25 July 2007 (UTC)

Thumperward

Like I've tried to tell you more than enough times, instead of simply reverting, try to actually improve an edit you disagree with. Right now, yes, the version isn't perfect (I'm comparing and contrasting your version, and will be fixing it up and incorporating things I think are better in your revert version), but your version is far worse. and the upper level could be supplied by the X Window System, barely makes sense to me, and I've been using computers far longer than I remember. Saying that it provides the upper level is inaccurate and confusing. How the heck is someone who knows little about computers (let's admit it, if you know nothing of them, you aren't reading this article) going to understand what "Upper Level" is? A "GUI environment", or something to that effect is much clearer, and to the point. device drivers and daemons had stalled and was incomplete. is grammar that just pains me. In addition, making specific sections for CLI and GUI are needed, as otherwise it gives the impression that the main user interface is the command line, which is very off-putting to anyone that would consider using Linux, and insulting to the grand majority of modern Linux users (ignoring servers) that use GUIs more than CLI. Separating in the history of linux section is needed to distinguish between what came before linux (which would be impossible to understand the beginnings of without), and what actually came with linux --L^augh! 14:13, 25 July 2007 (UTC)

This is just bat-signal level projection. By continuing to revert to a disagreeable version (which pushes the "filled a gap" POV that has already been discussed on here) you're not providing a solid foundation upon which to discuss more nuanced changes. I'm not going to discuss style minutae in a section which has just been rewritten to present an overall disagreeable POV. I have no problem discussing edits here, and most of the rewording of the last two edits is fine. But doing so while accusing me of being a troll and blanket-reverting my own edits is not going to get me in a good mood. The be bold, revert, discuss cycle favours working from the present version of an article to improve it by backporting good bits. Chris Cunningham 14:48, 25 July 2007 (UTC) Resolved. Apologies. Chris Cunningham 14:55, 25 July 2007 (UTC)

Stability

Okay, my bad. The current version is pretty decent, though I still disagree with the overuse of subheaders. We should be able to work from here. Chris Cunningham 14:53, 25 July 2007 (UTC)

It's both of our bads. Anyway, ideally, I'd like to condense that entire section into a few paragraphs. That really should be handled in the history of linux article, not here, but we do need to give a summary. The "naming" and "Milestones" sections can probably be killed with fire- they're somewhat opinionated no matter how you put it, and don't really add anything to someone reading this article. The other sections should be separated somehow, although if we can convert the first one into a single neat paragraph, just starting a new paragraph would be fine there. As far as the UI thing goes, I still think CLI and GUI both need their own section, but we should put a little bit of information on what a UI is too. Maybe a {{See also}} to User Interface, and a simple one or two sentence explanation of what a computer UI does? --L^augh! 15:07, 25 July 2007 (UTC)

The thing is, I see "no header" as just being another section, thus negating the need for one. As for the Milestones section, yeah, it needs totally restructured. It's utterly arbitrary and isn't chronological. More time for this later I hope. Chris Cunningham 15:19, 25 July 2007 (UTC)

Actually, should we knock out those two sections? They could be added in to the actual History section later if they're notable enough to have a place in this article (I personally don't think so, but "in 199x, Linux gained a large step forward with the release of OOo" in the history section makes a lot more sense than a "Milestones" subcategory. The naming controversy could probably be mentioned in it's own section, down near the bottom, so it could get a bit more care to help remove the POV-ness, without taking away much attention or adding WP:UNDUE to it. As far as UI goes, that's really personal opinion I suppose, the MOS isn't really clear. Judging by Today's Featured Article, I'd say that having a heading and then a subheading right next to each other is fine --L^augh! 15:36, 25 July 2007 (UTC)

I don't mind if the sections are removed. They were originally brought here in a large restructuring by User:Chris Pickett a while ago, by translating from the featured French article. I really think we should avoid overstretching ourselves by trying to cover the whole of the free software movement if possible. And yeah, the sections is no biggie, it just go caught in the way of the earlier reverts really. Chris Cunningham 20:24, 25 July 2007 (UTC)