Talk:Linux kernel/Archive 2

Kernel Version
This article is about the linux kernel, past, present, and future. I posted some useful info on how a person can know what version they are running. Chealer deleted it saying this article is not a usage manual. Knowing what kernel version you are using is not a usage manual. It is useful information that relates to the article. What do you think? Daniel.Cardenas 11:31, 23 August 2006 (UTC)
 * It all depends on the tone of the article. A descriptive article which indicates the differences between releases might be fine. A "how to find out what kernel you're using" piece less so. The article suffers for extlinkitis already anyway. Chris Cunningham 10:47, 26 September 2006 (UTC)

Mascot
Hi all,

I think that the mascot section should be either removed or considerably expanded. An extra sentance can be added to the top or something, but the current eight words doesn't seem to be enough for its own header. Ultra Loser 11:32, 11 September 2006 (UTC)


 * If anything, I think it should be removed. Tux the mascot doesn't really have much to do with information on the Linux kernel, it just seems to be extraneous information. tgok 15:48, 7 October 2006 (UTC)

Linux Kernel Implementations
Is it possible to create a new wikipedia page about Linux Kernel Implementations on multiple hardware architectures? This page will contain the list of "Linux Kernel portability" from this page.

oops
mention about Linux kernel oops (maybe in Technical features section)

Tagged lists
At least the "new features per release" section should be folded into the general history. I don't see that a list of every single architecture Linux has been ported to is very useful either, given that it's unlikely to ever be complete. Chris Cunningham 10:28, 11 November 2006 (UTC)

md
The MD disambiguation page points here to "Linux (kernel)", but this article never mentions "md". Should we mention "md" in this article, or should we make a "md (Linux)" article, or is there some other article that the MD disambiguation page should point to? --75.27.231.105 03:17, 15 November 2006 (UTC)

Translating "History of Linux"
Please note the translation effort currently at Wikipedia:WikiProject Linux/Translatation:Geschichte von Linux. - Samsara (talk • contribs) 15:23, 18 November 2006 (UTC)

Proposal for better disambiguation (see Talk:Linux (disambiguation))
I think GNU/Linux should be the main operating system page, Linux (kernel) should be about the kernel, GNU/Linux naming controversy should remain as is, and in fact Linux should be a disambiguation page that points to all three, saying this: Linux either refers to GNU/Linux the operating system, also known as just Linux, or Linux the kernel. For more information on this disambiguation, you can read about the GNU/Linux naming controversy. --Chris Pickett 04:33, 6 December 2006 (UTC)
 * I just discovered a 5th page, Linux (disambiguation). In opinion, this should be moved to Linux.  Please discuss at Talk:Linux (disambiguation).  --Chris Pickett 04:45, 6 December 2006 (UTC)
 * I agree. I think it'll be difficult convincing some people to change the name of the operating system page to GNU/Linux though. I find the current first paragraph of that page confusing, as GNU/Linux is far more common than any other operating system using the Linux kernel. Guyjohnston 01:19, 19 January 2007 (UTC)

Pronunciation (transclusion)
Hi, Linux kernel is now transcluded from, but the references don't show up at Linux kernel. Maybe we should copy the info back..? --Kjoonlee 16:55, 25 December 2006 (UTC)


 * Interesting. Looks to be a problem with the MediaWiki software. --Imroy 05:28, 26 December 2006 (UTC)

Pronunciation - contradiction?
It seems that the section on pronunciation is contradicted by http://www.youtube.com/watch?v=5IfHm6R5le0 : Linus uses [ˈlɪnʊks] in that video (or something close to it), rather than 'li' using [ee] as mentioned in the text. Has he changed his mind at some point? --Chris Wood 14:03, 5 January 2007 (UTC)
 * I've gone ahead and changed the section, referencing the Youtube video --Chris Wood 16:43, 20 January 2007 (UTC)

Seminar Paper On Linux Kernel 2.6
Hi,

I have added a new post " Seminar on Linux Kernel 2.6 " Hope It will be useful for Students who like

to study the basic concept of linux kernel 2.6.

Virtual11234

Poor Place for GPLv3 Discussion
Original paragraph:


 * Currently, Linux is licensed under version 2 of the GPL, and there is some controversy over how easily it could be changed to use later GPL versions such as the upcoming version 3 (and whether this is desirable).[13] Torvalds himself indicated in version 2.4.0 that his own code is only under version 2.[14] However, the terms of the GPL state that if no version is specified, then any version may be used, and Alan Cox pointed out that very few other Linux contributors have specified a particular version of the GPL.[15]

There are a number of problems with this paragraph. Foremost is that there are no reliable sources to back anything up; everything on this topic is just hearsay. When the two warring sides agree, or when a court case reaches a decision, then we'll have material worthy of Wikipedia. Right now, a simple "converting Linux to GPLv3 is expected to be quite difficult" (which most everyone seems to agree on) is about the only statement that's appropriate to include in Wikipedia.

Especially troubling is the clause: "However, the terms of the GPL state that if no version is specified, then any version may be used." This is provably false and probably a textbook example of wikiality. Read the GPLv2. Notice how the "or later" clause comes after "END OF TERMS AND CONDITIONS". That means it's NOT a part of the license. Has Linus ever included this clause in his code? Certainly not. Therefore, and this is provable, the "GPLv2 or later" suggestion does not apply to the Linux kernel. Here's more, from about as authoritative a source as you can find at the moment: http://lkml.org/lkml/2006/1/27/339

Please do not use Wikipedia to advance your own personal agenda. And please ensure that you have actually READ the license before stating something as fact. Thank you. -- Bronson


 * Hey, Bronson. Speaking of "personal agendas" and "actually READING licenses", better check yer facts next time:


 * "9. The Free Software Foundation may publish revised and/or new versions of the General Public License from time to time. Such new versions will be similar in spirit to the present version, but may differ in detail to address new problems or concerns. Each version is given a distinguishing version number.  If the Program specifies a version number of this License which applies to it and 'any later version', you have the option of following the terms and conditions either of that version or of any later version published by the Free Software Foundation.  If the Program does not specify a version number of this License, you may choose any version ever published by the Free Software Foundation."


 * 80.233.255.7 14:18, 6 February 2007 (UTC)


 * Hi Anonymous. Notice that the text you quoted comes after "END OF TERMS AND CONDITIONS."  That means it's not a part of the GPL, it's only a suggestion.  -- Bronson


 * I deliberately included the section number in the quote. Only terms and conditions are numbered in the GPL.  Your comment indicates that you have neither read the licence nor my reply.  80.233.255.7 23:52, 10 February 2007 (UTC)


 * I've reverted both of your edits. Also, the paragraph you modified cites the Linux kernel mailing list;  I wouldn't call that an "unreliable source".  80.233.255.7 14:30, 6 February 2007 (UTC)


 * Well, I'm not going to get into an edit war over something like this. I truly have no agenda, other than keeping bias out of Wikipedia.  Here is my paragraph:


 * The Linux kernel is licensed under version 2 of the GPL. The FSF is currently working on GPL version 3 to try to clarify the license terms and close some loop holes. It's not clear yet how realistic a project it would be to move the entire kernel to GPLv3. And, because V3 is slightly less permissive than V2, there is debate within the kernel community as to whether it's even desirable.


 * All I ask is, which one is more appropriate for inclusion in an encyclopedia? Answer: probably neither since this information is just too new.  Also, you might want to reexamine your idea of authoritative source -- it does not agree with Reliable sources.


 * This paragraph is way better than the one I reverted. Some references and I'll be happy to see it in the article.


 * I think it's safe to say that most news sites pick up posts from the Linux kernel mailing list for their stories. I don't see how is such a news site more reliable than the mailing list.  80.233.255.7 23:52, 10 February 2007 (UTC)

Alan Cox not mentioned?
Alan Cox's involvement isn't mentioned anywhere at all in this article. Seems to be an oversight, but I don't have much knowledge to add. Maybe others can. Gronky 19:31, 4 February 2007 (UTC)
 * And so Andrew Morton, Ted Tso, Ingo Molnar, Andi Kleen, Al Viro, David Miller .. (list can never end). It is impossible to even only mention the core kernel group since they are many. Darwish07 (talk) 18:08, 23 January 2008 (UTC)

Linux supports such and such architectures

 * from User talk:ChrisRuvolo

Hi. It bothers me that you did not explain your reasoning when you reverted my edit to Linux kernel article. I would appreciate if you did that at the article's talk page. Thanks. 80.233.255.7 21:14, 14 February 2007 (UTC)


 * Hi. Showing the breadth of Linux portability is important. Removing the entire list with the comment "The kernel probably has its own documentation" doesn't indicate any reasons why this information would not be encylopedic.  Is the information lengthy?  Yes, but that is okay.  Wikipedia is not a paper encyclopedia.  --ChrisRuvolo (t) 21:38, 14 February 2007 (UTC)


 * Thank you.


 * I entirely agree that it's important to emphasize portability. I disagree with the approach.  The list can be summed up in a sentence "Linux supports many architectures" followed by a link to the section in kernel docs where the architectures are listed.  Because Wikipedia is also not an indiscriminate collection of information, plus there's this "unencyclopedic lists" template at the top of the article.  80.233.255.7 22:29, 14 February 2007 (UTC)


 * I find this material encyclopedic, but if people find it distracting in this article, I would suggest that it be moved into another article, using summary style. Perhaps Linux kernel portability or List of architectures supported by the Linux kernel.  Your thoughts?  --ChrisRuvolo (t) 13:45, 15 February 2007 (UTC)


 * Not sure what the policies have to say about "list articles", but I've seen a couple of those before, so maybe that would be an appropriate solution. 80.233.255.7 11:39, 16 February 2007 (UTC)


 * Let make the Linux kernel portability and supported architectures so we can lose the Laundry List tag; all that detail is rather distracting in the main article. Intersofia 13:30, 6 April 2007 (UTC)

GNU paragraph in intro
GNU has nothing to do with the original release of Linux. Torvalds specifically differentiated Linux from GNU in the initial kernel announcement. I'm happy for this to go in the history section itself, but putting it into the intro is attempting to force GNU into the whole "operating system" thing. This isn't the OS article, it's the kernel article, and GNU is only important insofar as Torvalds may not have bothered had GNU a kernel in 1991. Removing this again. Chris Cunningham 23:59, 31 March 2007 (UTC)


 * GNU has very much relevance to why the Linux kernel was important. Removing the paragraph is in my opinion absurd. --SLi 11:13, 2 April 2007 (UTC)


 * What's the justification? Again, this isn't the Linux article. This is specific to the kernel. Chris Cunningham 11:23, 2 April 2007 (UTC)


 * Well, the justification is, as clearly was stated in the text you removed, that it was relevant to the Linux kernel importance. --SLi 23:50, 6 April 2007 (UTC)

History Deletions
What is the purpose of these deletions from the History section ? "At the time, the GNU Project had created many of the components required for a free software operating system, but its own kernel, GNU Hurd, was incomplete and unavailable. The BSD operating system had not yet freed itself from legal encumbrances. This left a space for the Linux kernel to fill, and despite the limited functionality of the early versions it rapidly accumulated developers and users." is an important part of the history IMHO (and it seems at least 3 other users) - Rythie 19:14, 2 April 2007 (UTC)


 * I keep hearing this. I'm yet to hear why it is important, expect in some abstract sense where all free software should be promoted everywhere, and "isn't it lovely that a free OS was in development" yadda yadda. The Linux kernel wasn't developed to fill in the missing gap in the GNU project. That it later became the de facto GNU kernel for approximately everyone in the community is an important note for the Linux article, which discusses the larger aspect of actually using the kernel to do some work, but it's not so directly relevant to the history of the kernel that it deserves a place in the intro. Chris Cunningham 19:53, 2 April 2007 (UTC)


 * Glyn_Moody's book 'Rebel Code: Linux and the Open Source Revolution' which covers the birth of the Linux kernel amongst other things. The book talks about 386BSD in quite a bit of detail and the fact that Linus was aware of it at the time on p65-67 and Linus is quoted as saying "if 386BSD come out a year earlier, I probably wouldn't have started Linux". Hurd is barely mentioned except to quote one of Linus's initial usenet postings on p45 . Rythie 23:15, 3 April 2007 (UTC)


 * Okay, but that's not really the focus of the section I removed, which insinuated that Linux was written to solve GNU's kernel problem. If you feel like re-adding this without the revisionism go right ahead. Chris Cunningham 08:28, 4 April 2007 (UTC)


 * That's not at all what the section says. It says: "This left a space for the Linux kernel to fill". Nowhere in this does it imply that Linux was specifically developed to fill it, it only says that it filled a gap. I think it's rather delusional to dispute it did. Is it your position that the availability of the GNU userland in no way affected the popularity of Linux (the kernel)? --SLi 00:10, 7 April 2007 (UTC)


 * Delusional? No, I don't think it's "delusional", thanks. And that's a straw man, I didn't imply there being no link at all; I just didn't like the implication presented by such a section making up a large part of the intro. Chris Cunningham 12:25, 7 April 2007 (UTC)


 * I was thinking that this background around BSD and Hurd is not mentioned in History of the Linux kernel, so it would perhaps better to put that detail in there instead of mentioning it here --Rythie 10:23, 7 April 2007 (UTC)


 * Indeed. As I said before, this is fine to note in the kernel's history (both here and in the main article), but the previous wording and positioning was somewhat questionable. Chris Cunningham 12:25, 7 April 2007 (UTC)


 * Do you mean that this article, "Linux kernel", should not talk anything about the history of the Linux kernel? And if you reread what I said, it was not a strawman, but a question like "this is what you claim - I believe it's false, unless X. Do you claim X?". And contrary to what you say, the wording doesn't suggest _at_all_ that Linux was written to fill that gap. --SLi 13:02, 7 April 2007 (UTC)


 * No, this wasn't what was suggested at all. And it was a straw man; you're asking me to defend a mischaracterisation of my argument (that "he availability of the GNU userland in no way affected the popularity of Linux": of course it did, but the two aren't so interlinked regarding the kernel's development as to warrant a large part of the intro discussing the history of GNU). Chris Cunningham 14:35, 7 April 2007 (UTC)


 * In my opinion its placing is also appropriate. It's indeed THE piece of history that made the rapid development of Linux kernel possible. --SLi 13:23, 7 April 2007 (UTC)


 * If you still feel it's wrong for it to be there, perhaps we should make a Request for Comment. My position is that you started the edit war by removing a very central piece of information that had been in the article since 2006 as "utter irrelevance", and persisting in reverting attempts to put it back by me and another person that just saw your removal as absurd (I think claiming it's irrelevant is revisionism if anything). --SLi 13:32, 7 April 2007 (UTC)


 * Could you calm down with the personal commentary? And I'm still not seeing any actual evidence which suggests that GNU was axiomatically important for kernel development in a way which necessitates bringing this up so strongly in the intro. This can certainly be argued for things like gcc, but not the userspace. And further, this doesn't defend the old wording, which gave a whole paragraph of the intro over to the development of GNU rather than the Linux kernel. Chris Cunningham 14:35, 7 April 2007 (UTC)

References added to the current dispute
I have no idea what the first refs is purported to be, but it doesn't support the given argument at all. The second provides valuable commentary and should be integrated properly rather than tacked on to a paragraph. The third likewise has nothing to do with the given argument. In the interests of not continuing an edit war, any chance on working on this? Chris Cunningham 14:41, 7 April 2007 (UTC)


 * I think trying to work on this is much better than an edit war. I appreciate that. Thanks :)
 * Hmm. From the first ref:
 * However, it took until 1991 for Mach 3.0 to be released under a free software license. In the the same year Linus Torvalds started to fill the kernel gap in the GNU system by writing the Linux kernel.
 * Ok, I agree this is poorer than the contested text in the article, since this gives the impression that Linus specifically wanted to fill that gap.
 * The third ref:
 * They had written almost an entire operating system by the early 1990s, but the kernel was missing. Fortunately, Linux appeared to fill this gap.
 * This I believe is more neutral and true than that in the first ref.
 * My argument is that while Linux was not written to fill a gap, it did fill a gap, and I believe that's what the contested paragraph says:
 *  This left a space for the Linux kernel to fill, and despite the limited functionality of the early versions it rapidly accumulated developers and users.
 * Sure if you think the paragraph wrongly gives the impression that Linux was specifically written to fill that gap it could be worked to deny that (while that's not how I read it).
 * Would you find changing the above sentence to "While Linux was not specifically written for the GNU system, this left a space for the Linux kernel to fill in it, and despite the limited functionality of the early versions it rapidly accumulated developers and users." acceptable? --SLi 14:54, 7 April 2007 (UTC)


 * The point is that the second ref goes at this from the Linux angle, which I believe is the correct approach, rather than the GNU angle pushed by the FSF. The second ref specifically points out that Torvalds wanted to create his own OS and GNU code was the quickest way of getting there; that the "gap" as such came in the form of the userspace utilities, a free compiler and so on. As this is the Linux article and not the GNU one, this should be the way the section is phrased. Chris Cunningham 15:35, 7 April 2007 (UTC)


 * I've attempted a rewrite to address this, shedding the two refs which present the "fill the gap" frame from the POV of the FSF. Any better? Chris Cunningham 15:43, 7 April 2007 (UTC)


 * If you're satisfied with that, I'm very satisfied. I think I'm starting to see how you view this, and the part of the thing you disagreed with is something that I didn't consider important at all. The important thing to me is just that this piece of history is there, the fact that the kernel got developers because it and the GNU userland nicely complemented each other. I consider the text after your modifications at least not any worse than the original. :) Thanks. --SLi 15:58, 7 April 2007 (UTC)


 * Excellent. I'm quite pleased with the way this has worked out, so I've edited the same section in Linux to take a similar approach. Chris Cunningham 16:33, 7 April 2007 (UTC)

Creator
I think it is necessary to change the note about Torvalds being Linux's "creator" and instead call him the "principal author". Can we make this edit? Dylan Knight Rogers 14:07, 10 June 2007 (UTC)


 * Errr, why? This would be significantly less accurate. It is indisputable that Linus is the kernel's creator. It is highly disputable that he is the principle author at this point. Chris Cunningham 13:07, 12 June 2007 (UTC)


 * it's not that big a deal, perhaps, but with the whole planet involved in wikipedia; a certain attention to detail amongst those who have the time and motivation is probably a good thing: the most accurate description of the situation that I've heard is that Thorvald is the person around which linux grew -  along with that, he is the namesake, the person who concieved of and launched the effort, the principle coder in the earlier versions, and that he continues on as a symbol and as benevolent dictator for life.--66.245.28.149 20:48, 7 July 2007 (UTC)

Binary Fission and Cell Differentiation
The reader would be served by splitting this article into two articles: one more technical, one less technical.

For example:


 * Pronunciation and SCO controvery would go into the less technical article.
 * The fact that the kernel is written in C, would go in the more technical article.

maybe even three articles - based on the concern of the reader: technical, legal, other (including history)

--66.245.28.149 21:27, 7 July 2007 (UTC)

Stable Version History
Maybe it's just me, but it looks like the later versions have too many features listed. I think the focus here should be on listing the major features - like support for a new (major) architectures, SMP support, etc. Fry-kun 02:07, 10 July 2007 (UTC)


 * Welcome to Wikipedia. :) Adding trivial factoids to large, indiscriminate lists of information is a very easy way to contribute, which is why they always get unbalanced and out of control. Chris Cunningham 10:06, 10 July 2007 (UTC)

Capturing 2.4 maintenance
The 2.4 kernel is still being maintained with the most recent version 2.4.35 being announce on the 26th July 2007 by kernel maintainer Willy Tarreau ref here. Where do we put this ? Could we extend the infobox to have a "Maintenance release:" or do we just have it in the main text ? Ttiotsw 09:06, 28 July 2007 (UTC)

PowerTOP
Sent here from Drawing_board. PowerTOP is a Linux utility, mainly for Intel CPUs, released by Intel and available at http://www.linuxpowertop.org/. If you search for “powertop” at Google or Clusty, you will find mentions of the utility first. OTOH, if you search Wikipedia, you'll find completely different stuff:, and not a mention of PowerTOP.

Question: is PowerTOP notable enough to have its own article, or should it just have a disambiguation page? --AVRS 11:38, 21 August 2007 (UTC)

menuconfig reference
89.220.88.209 17:37, 14 October 2007 (UTC) What is that reference to menuconfig doing there?

patents
what parts of the linux kernel violate microsofts patents. COuld we be sued--Tuxthepenguin933 00:56, 12 November 2007 (UTC)
 * Most likely none, so/because Microsoft won’t name any, just spreading FUD. This page is not for discussion of that, though, only for discussion of the article. Please, use one of the forums out there. --AVRS 13:45, 12 November 2007 (UTC)

kernel architecture
hey guys...am doing my final yr Comp engg., i want to know the architecture of linux kernel,hw it works n the best way 2 understand it....... can any1 here plzzzzzzz help me with some ideas n reffernce.... fleming_kd@yahoo.com plz mail ur views to dis guys....plz guys help me....waitig 4 ur mails... —Preceding unsigned comment added by Fleming kd (talk • contribs) 10:29, 22 January 2008 (UTC)
 * NOT a help desk :). Please ask your technical questions in a related technical forum. Darwish07 (talk) 12:30, 23 January 2008 (UTC)

Performance claims
At http://en.wikipedia.org/w/index.php?title=Linux_kernel&diff=185078477&oldid=184852434 a statement that programs run on Linux 5-7 times faster is introduced. There is no citation for this, this is isn't what my experience tells me, and nowhere on the web is such a claim sustained. I propose reformulating the statement, I don't think we will find a citation for something like that.--Alecsescu (talk) 12:35, 23 January 2008 (UTC) Ok, I did a primitive deletion of the statement.--Alecsescu (talk) 12:39, 23 January 2008 (UTC)
 * I agree. Also it's not the place of a Linux Kernel article to discuss _very generally_ the performance of user-space applications. This depends on many different factors including compilers, libc, and other core libraries performance. Darwish07 (talk) 13:59, 23 January 2008 (UTC)

Changing the article logo (No Kernel messages displayed)
Looking at the article logo, there's no single Linux Kernel message in it. It only displays some of Knoppix user-space messages. I think to make the screenshot more related, it needs to capture the early messages of Linux kernel boot (Linux Kernel v2.6.21-rcX bla bla) till init starts or till some of meaningful information about the running kernel is displayed.

It would also be much more better if it's a stock kernel (i.e. not a distribution customized kernel).

Anyone who have a suitable image/screenshot ? -- Darwish07 (talk) 18:20, 23 January 2008 (UTC)

Refactoring with Linux
I think this article and Linux could use some refactoring. I discuss it in more length at Talk:Linux. Feel free to weigh in there. Thanks, WalterGR (talk | contributions) 19:01, 29 March 2008 (UTC)

Non-GNU GPL Linux kernel
I was reading the Wikipedia Signpost today, and it said "Merkey offered $50,000 for a copy of the Linux kernel that was not licensed under the GNU General Public License". I thought all Linux kernels were released under the GNU General Public License. The article states: "The Linux kernel is released under the GNU General Public License version 2". I have not been able to find anything that suggest that there are copies of the Linux kernel not licensed under the GNU GPL in the article. Could someone please explain? Joshua Issac (talk) 13:54, 18 June 2008 (UTC)


 * They are. Merkey's request was preposterous, and was treated as such by most of the community at the time. Chris Cunningham (not at work) - talk 14:20, 18 June 2008 (UTC)

Refs & External links
The refs on this page are not well organized at all. The external links on this page are ridiculous. I'm going to toss a number of ELs, posting the ones I remove here so nothing is immediately lost in case there are questions about their value. Chaldor (talk) 20:35, 20 September 2008 (UTC)

How big is it?
Could we include info about how how much space the kernel uses on a disk? just curiosity. ubuntu has a bunch of versions of the kernel on my disk and every time i download a new one i wonder if it takes up a noticable amount of space, but then to consider that the kernel is in so many devices it must be very small.217.149.150.2 (talk) 06:27, 18 October 2008 (UTC)
 * the openwrt kernel has support for multiple wireless cards plus the basics and is 4 megabytes. most of that space is drivers and some is to make it run faster.Scientus (talk) 14:24, 26 December 2008 (UTC)
 * As of 2.6.28 the kernel source is about 50 megabytes of bzipped tarball, uncompressed it will obviously be more. The main kernel itself is typically of the order of a few megabytes depending on the options used when compiling it. The modules shipped with a distro kernel (which tends to have virtually every driver in the tree compiled as a module) however are pretty huge. On my debian system the /lib/modules/2.6.26-1-amd64/ directory comes out at 79MB (that may include the odd out of tree module but i'm pretty sure the bulk of it is stuff that shipped in the kernel package). Plugwash (talk) 19:23, 23 January 2009 (UTC)