Talk:Linux/Archive 43

Latest comment: 10 years ago by Dsimic in topic GNU Bash
Archive 40Archive 41Archive 42Archive 43Archive 44Archive 45Archive 50

GNU and me too

The lead pgraph says The defining component of Linux is the Linux kernel, an operating system kernel [...], also misread "The defining component of an OS is the kernel". This page is "Linux" the computer OS. Let the page Linux kernel say "The Linux kernel is a Unix-like operating system kernel". Also it is misread "The Linux kernel is the Linux kernel" or "The Linux distribution is defined by the kernel", both of which are misleading.

For a lead? For a grandmother hearing "Linux", Linux should mean the entire distribution, (not just the kernel), because it will mean what she can have instead of Windows XP. "Linux kernel" is used elsewhere, under the lead, in the body. Proposed:

A Linux distribution is a software stack configured through a Linux kernel.[1] A distribution of Linux is largely documentated and support by the GNU project. .

It is not for nothing that the title GNU/Linux is an immortal redirect title to this article. GNU should be mentioned definitively. Thanks. — CpiralCpiral 23:09, 16 April 2014 (UTC)

  • Oppose. I won't revert you again – I'll leave that someone else, confident there probably will be someone else – but I really don't think posting this cryptic proposal (which, btw, doesn't match what you actually did), waiting 2 days and getting no response constitutes discussion indicating you have consensus support. Msnicki (talk) 02:05, 19 April 2014 (UTC)
  • Comment - I've reverted the edit for a few reasons. As I mentioned in my edit summary, "A distribution of Linux is largely documented and support by the GNU project." makes no sense. I also disagree with Cpiral's assessment of how the current wording reads; I'm not seeing how it could be construed to read "The Linux kernel is the Linux kernel." I'm also not sure what Cpiral is trying to say about the redirect; it is [WP:NOTFINISH|neither immortal]] nor is it a "title", it's a redirect. - Aoidh (talk) 02:43, 19 April 2014 (UTC)
    • Is a distribution supported? Documented? If so then a distribution is documented and supported by some project.
    • If, as it now says, The defining component of Linux is the kernel [the Linux kernel], then it logically follows that "The defining component of Linux is the Linux Kernel". But the metalanguage of this article is about a distribution, is about a computer system, as defined by the analysis of the situation into hardware, user, and default software between the two. Thus the proposed "Linux is defined by a software stack configured through a Linux kernel" and "Linux is largely documented and support by the GNU project."
    • The search results page specialized search box says "There is a page named "GNU/Linux" on Wikipedia". To the search engine, if not the navigation command from a user who types "GNU/Linux", "GNU/Linux" is a title of a page on Wikipedia.
There's Linux as a "look and feel". Opposed there's Linux as a developer's or hobbyists kernel-modifications. The two are somewhat opposed as defines an audience.
The defining component of Linux, our lead now says, is the kernel. Well it depends on who is doing the component analysis. It depends on the audience including the GNU/Linux seekers. To GNU, and to me, the components are the computer, the user, and the look and feel of the system between them. The key component here, per my audience-analysis, being "the system". The audience for this article is the user, and the subject for this article is "the system". The audience is not interested in how a kernel defines Linux.

What defines Linux for our audience?

  • The lead depicts Linux running on a supercomputer. (I think supercomputers are defined elsewhere by key hardware components.) "Linux runs supercomputers." Not bad, sounds great. Who cares what version and kernel mods might have taken place when all they need to know is that Linux runs supercomputers. Another OS or kernel could have been be modified to run that supercomputer, but Linux runs supercomputers.
  • The first sentence says Linux is a computer operating system. But does that three word phrase describe the subject of this article Linux (GNU/Linux)? I read that phrase more as a "computer system". Is the subject of our Linux introduction going to mention some "operating system" feature? (The GNU C Library, as the kernel, could mislead our Linux audience.) The "operating system" interpretation of computer operating system could define Linux by the kernel if the audience was OS hobbyists and OS developers. Let's line vote by outline.
  • The TOC as an outline says sections exist for "creation" and "development", and "programming" of Linux, scoring three points for the defining component of Linux being "the operating system kernel", for our hobbyists and developers.
  • Scoring ten points is the user-definition of Linux as a computer system in its entirety, a "look and feel" of a computer system design.
    1. 1.3 Commercial and popular uptake (user uptake)
    2. 2.1 Design of User interface
    3. 2.2 Design of video input (user input)
    4. 4.1 Desktop uses (user)
    5. 4.1.1 Performance and applications (user's look and feel)
    6. 4.1.2 Components and installation (installation as user experience)
    7. 4.2 Servers, mainframes and supercomputers (look and feel)
    8. 4.3 Embedded devices (content is mostly look and feel)
    9. 5 Market share and uptake (executive summary, look and feel)
    10. 6 Copyright, trademark, and naming (popular user uptake due to look and feel),

The "computer system" reading should be defined by the look and feel experienced by its users and onlookers. Finally, "computer operating system" here does by my count mean "computer system" is more a primary topic than "operating system". An operating-system orientation would define Linux the way the reader of Operating system would like it, by then reading Linux kernel. — CpiralCpiral 06:01, 19 April 2014 (UTC)

This is going to sound rough, so I apologize in advance, but anyway – if you don't care about how and why Linux runs supercomputers, and what's under the hood of an average desktop, why should I care about your essay promoting managerial and GNU buzztalk? — Dsimic (talk | contribs) 13:26, 22 April 2014 (UTC)
Because I care. — CpiralCpiral 01:02, 24 April 2014 (UTC)

What is "Linux", when it is the single worded question "Linux?" typed into the Wikipedia search engine? The answer could be "a key component of an operating system", but I'm raising a disambiguation issue against it. I'm trying to make an objective, impersonal, appeal to a move from

Linux
  1. Operating system kernel
  2. Operating system distribution

to

Linux
  1. Operating system distribution
  2. Operating system kernel

in the lead paragraph. Blame the new hovercard feature :)

I see two contexts of "Linux" the word—kernel context and distro context. I can simply count occurrences and divide, into one side or the other, 1) the sections of our article, 2) the 239 Linux-in-title articles on our wiki. Intuitively I'm thinking that the reality of "Linux" in the English language is on the "distro" definition, not the "kernel" definition, and so in the lead paragraph (that fits a hovercard), that we should mention, via the term "GNU project", the popular Free Software Foundation giant and deemphasize the esoteric things that most people couldn't nod to, like the technical meaning of an "operating system" verses the commonly understood meaning of an "operating system". — CpiralCpiral 01:02, 24 April 2014 (UTC)

Do you really think that the majority of regular people could nod to the term "GNU project"? Joe Average doesn't give a fu*k about Linux, GNU or whatever makes up some... Linux? A distribution? Can I have my Outlook calendar on that? Whaaat, I can't? Who made that unusable thing?
Wake up. We're just spending our energy over things that 95% of people neither care nor know about, instead of doing something more productive. — Dsimic (talk | contribs) 02:47, 24 April 2014 (UTC)

Release date.

So according to this article, linux was released in 1991, 23 years ago, but on the Linux Kernel article it states the Linux Kernel initial release was 22 years ago. I'm guessing that's because the Linux Kernel page states September as the month, and this page just goes by the year, but it seems strange to have a page about the operating system, stating that the operating system came out a year before the kernel was released. Lmcgregoruk (talk) 16:01, 12 May 2014 (UTC)

Use DMY or MDY?

I was going to change eg. "February 8". See "first released on 5 October 1991" in lead. Then saw template-"Use mdy dates|date=November 2012". Didn't bother to find first use. I think this should be other way around and see my recent edites in History of Linux. Would be best if these article where in sync. And refs in sync with main article. At least not all three formats. YMD a comromise there? comp.arch (talk) 13:22, 19 May 2014 (UTC)

Operating System versus Kernel

That a difference exists between a "kernel" and an Operating System seems technically inaccurate. That difference doesn't seem to exist. See what the USPTO says the name Linux represents:

Go to the US Patent and Trademark website, here:

USPTO Trademark Search.  

Select "Basic Word Mark Search.

Enter "linus torvalds" into the search field and and change the value of Field to "Owner Name and Address".

Click "search". Select the result that showing as "Live", then click the "TSDR" button.

The following text is from the Status tab, under "For", "computer operating system software to facilitate computer use and operation".

A shell isn't part of the kernel but the command-line is the interface between user and OS. Isn't the shell (e.g. bash) also viewed as the interface to the kernel?

Kernel.package (talk) 04:14, 18 May 2014 (UTC)

The difference between a kernel and OS in general is computer science and some source for that could be found I'm sure. Kernel has a meaning (that can be stretched). OS would include the kernel and more (how much more, is a little fuzzy..). What the USPTO says (or Linus in his application) would not define either in general or for "Linux". Linus speaks of his kernel as a "monolithic kernel", admitting it is a kernel. I'm sure he would agree that an OS is more than the kernel, but he happened to apply for a trademark on the OS (as a whole) not the kernel. What I find more interesting are his dead trademarks ("LINUX COMPATIBLE")..:
Word Mark LINUX
Goods and Services IC 009. US 021 023 026 036 038. G & S: computer operating system software to facilitate computer use and operation. FIRST USE: 19940802. FIRST USE IN COMMERCE: 19940802
Mark Drawing Code (1) TYPED DRAWING
Word Mark LINUX COMPATIBLE
Goods and Services (ABANDONED) IC 009. US 021 023 026 036 038. G & S: Computer hardware
Mark Drawing Code (3) DESIGN PLUS WORDS, LETTERS, AND/OR NUMBERS
Design Search Code 03.15.15 - Penguins; Puffins
03.15.26 - Costumed birds and bats and those with human attributes
Word Mark LINUX COMPATIBLE
Goods and Services (ABANDONED) IC 009. US 021 023 026 036 038. G & S: Computer hardware
Mark Drawing Code (1) TYPED DRAWING
comp.arch (talk) 13:57, 19 May 2014 (UTC)
"Linus speaks of his kernel as a "monolithic kernel", admitting it is a kernel. I'm sure he would agree that an OS is more than the kernel..."
He does, actually:
"Sadly, a kernel by itself gets you nowhere. To get a working system you
need a shell, compilers, a library etc. These are separate parts and may
be under a stricter (or even looser) copyright. Most of the tools used
with linux are GNU software and are under the GNU copyleft. These tools
aren't in the distribution - ask me (or GNU) for more info."
https://www.kernel.org/pub/linux/kernel/Historic/old-versions/RELNOTES-0.01
189.127.212.79 (talk) 00:07, 24 May 2014 (UTC)
The kernel of an OS is the component of the OS that does the resource management. An OS also includes many components shipped with the kernel in a system installation(but not necessarily things the users install after that), ranging from libraries like glibc to system daemons like systemd, PulseAudio and a display server. An OS has 3 main functions: resource management, providing runtime environments for applications, and providing a UI for user interaction. The kernel implements the first one, and may take certain important roles in the latter two, but the latter two need extra programs to actually be implemented. Some people even classify core application software shipped with the OS as part of the OS.
Linux is not a single operating system, it's a family of operating systems all using the Linux kernel, and commonly also glibc(but not necessarily). Each distribution should be considered to be a distinct OS, or at least a variant of a certain form of linux. For example, Debian and Fedora should at least be a variant of "desktop Linux" while the so-called desktop Linux is essentially a very different operating system from those that don't include GNU userland, like Android. Therefore, I propose that the lead be edited to read: "Linux is a family of Unix-like and mostly POSIX-compliant computer operating systems assembled under the model of free and open source software development and distribution. The defining component of Linux is the Linux kernel, an operating system kernel first released on 5 October 1991 by Linus Torvalds. The Free Software Foundation uses the name GNU/Linux to refer to "traditional" Linux operating systems that include GNU programs and libraries as a core component, which has led to some controversy."
This way it is possible to avoid two problems:
  • confusion of "traditional" (or GNU/)Linux with Linux in general, which may differ a lot
  • the view of Linux as a single operating system, instead of a group of them using the same kernel but possibly completely different userland
Busukxuan (talk) 08:19, 13 June 2014 (UTC)

Ported to more platforms than any other operating system

I'm not normally one to comment in talk sections like this, but if ever there were a statement that could use a big old "Citation Needed" link, this would be one. As best as I can find, NetBSD supports more platforms than Linux. I am not a NetBSD user, and I am a Linux user. Regardless, if the statement is correct, it should include some citations. If it is not correct, then perhaps a simple rephrasing to the effect of "been ported to more hardware platforms than most other operating systems" or "been ported to many additional hardware platforms". Just thoughts. ScottDaleRobison (talk) 21:48, 3 June 2014 (UTC)

Put up a {{citation needed}}. As for comparing to NetBSD, simple counting isn't going to work (aside from it being OR), because what constitutes a platform is a matter of definition. E.g. NetBSD has separate "ports" for mac68k and sun3, both Motorola 68000 series machines. The Linux kernel people never, AFAIK, bothered to formalize such a notion; if you look at linux/arch, they have a directory for m68k, with mac and sun3 inside it. Is m68k the platform? Is it mac or sun3? (Note that Linux/m68k has common code per processor type in linux/arch/m68k/platform; note that NetBSD initially considered acorn32, cats, and shark to be one arm32 port, but then split that in three.) QVVERTYVS (hm?) 23:44, 16 June 2014 (UTC)

Consensus?

Hello. I have added to this article a template warning readers about the POV dispute of the title, which was reverted minutes after the event by “Aoidh”. As you can see in the link, I explained that the addition of the template is supported by the very lengthy talk page (With its corresponding archives). He claims that consensus has been reached, but I don't see that claim supported by this talk page, quite the contrary, I see that no consensus has been reached and that instead the article title has on staying with the current title, in other words, inertia rather than consensus can be attributed the title of “Linux” to name this article. You can even find that the dispute continues (There has been a discussion in this month). Regards, QrTTf7fH (talk) 22:24, 29 July 2014 (UTC).

There is an overwhelming consensus that the article's title meets Wikipedia's policy on titles. You are welcome to go through the discussion and see this. A few editors creating accounts simply to come to the talk page and give their personal opinion of why GNU/Linux is the "correct" title and that saying otherwise is wrong does not create a lack of consensus. It is not a matter of numbers or a certain number of editors wanting to change the title, consensus is not a vote. Rather, consensus is ascertained by the quality of the arguments given on the various sides of an issue, as viewed through the lens of Wikipedia policy. The edit summary you used to include the tag noted that readers ought to know of this and other POV disputes however, the lede paragraph makes it unambiguously clear that there is a dispute about the name. As far as this article's title, however, I reverted the tag because there is no ongoing dispute but rather a continuous series of new editors wanting to change the article's title to fit a minority POV, and subsequent discussions explaining why that title is inappropriate for this article because when viewed in the context of Wikipedia's policy on article titles (which is how a consensus is determined), there has consistently been a consensus for the current title and no indication that anything about that has changed. That is why I reverted the tag. - Aoidh (talk) 22:38, 29 July 2014 (UTC)
Hello. You state (In other words) that disputes are not solved through democratic voting in Wikipedia, then it's irrelevant whether it's (In your view) a “few editors” which support naming this article GNU/Linux. You stated that consensus is “ascertained by the quality of the arguments”, and as a consequence, it's irrelevant whether the aforesaid “few editors” created their accounts “simply to come to the talk page and give their personal opinion”, since the history of the users who put forward an argument is not an element in ascertaining consensus. Could you please avoid adding irrelevant information to this discussion? Thanks you. For instance, in consideration to you and other parties interested in this discussion I'm not going to demand a proof for your claims that users who support naming this article GNU/Linux are “few” and created their accounts with the sole purpose of entering the discussion because we agree on the irrelevance of this point.
You said that there's an “overwhelming consensus”, but you're not providing a way to verify this, nor you stated in which way the supposed consensus is “overwhelming”. You're inviting me to “go through the discussion and see this”. I did read carefully a part of the discussion before making the change that you reverted, and glanced through the rest, and saw no signs of consensus. For every claim that was made in favor of Linux appealing to Wikipedia policy, a counter-argument was put forward. The Wikipedia policy don't speaks by itself, it can only be applied through interpretation. Your comment seems to imply that you have the capacity to know “the quality of the arguments given on the various sides of an issue, as viewed through the lens of Wikipedia policy” in an unequivocal way (Possibly by your own judgment, but maybe by a gold standard, of which none exists as far as I know).
I hold my argument that there's no consensus, given that claims in support of naming this article “Linux” were argued without either side accepting the other side's argument as valid, and both sides referred to the relevant Wikipedia policy.
Regards, QrTTf7fH (talk) 23:35, 29 July 2014 (UTC).
This archived discussion at Talk:Linux/Archive 41 § Page move: GNU/Linux should answer all your concerns about whether there's an existing consensus. Notice the discussion went on for a full year. There's a link to this in the "Archives" box right near the top of the page. Msnicki (talk) 23:46, 29 July 2014 (UTC)
I have already read that archive page, it does indeed make it clear than there's no consensus as for the name of the article, which is what I hold, and the reason to put the corresponding article tag. Note that “there's no consensus on moving the article to GNU/Linux” is not the same as “there's consensus in naming the article Linux”. Paraphrasing what I already said, the official Wikipedia policy states that when there's no consensus in the title of articles, the long-standing article title is kept (“Linux” in this case). QrTTf7fH (talk) 01:21, 30 July 2014 (UTC).

GNU Bash

Hello. Dsimic modified the caption of GNU Bash in this article. First he removed the mention of GNU altogether, then he switched the role of GNU to “developer” in the caption. It's a verifiable fact that Bash is part of the GNU project and it is its shell (I included a reference –note that self-published sources are considered reliable about claims on themselves–). I don't see how removing this fact improves the article; it may lead readers to think that Bash is an independent part of the operating system (GNU/Linux, called “Linux” in Wikipedia) while in fact it's a component of a larger project (the GNU project). These are the reasons of why I restored the previous version of the caption. Regards. QrTTf7fH (talk) 15:58, 1 August 2014 (UTC).

And I've reverted your edit for the reasons I explained in the edit summary; bash is not only used by GNU, so to word the caption in that way trivializes GNU's work; it is used by default on Linux, and OS X, and can be used on many other systems. That's a lot more than being "GNU's bash". That's why I reverted it, because it doesn't do a proper job of summarizing Bash and it's nature within system other than the GNU project. That it's part of the GNU project is in the bash article which is why wikilinks exists, but within the context of this article, the wording you're using is better used in the parent article (which it is). - Aoidh (talk) 19:51, 1 August 2014 (UTC)
Hello. Your claim is not supported by the article text. I left a message in my edit summary which reads: “The caption reads "Bash, GNU's shell widely used in Linux"; nowhere it says that only GNU uses it. Free feel to use the talk page.” QrTTf7fH (talk) 20:11, 1 August 2014 (UTC).
Yes, and telling people to "feel free to use the talk page" after they have already left a message there isn't helping. If this image caption were on the GNU article, that would make it relevant, but as explained, context is important and in the context of this article the caption you're trying to add isn't nearly relevant enough, and it also adds ambiguity in its meaning and and makes it seem like Bash is only for GNU; this is not the case. The software is developed by GNU; that is supported by the article text and saying "your claim is not supported by the article text" without further elaboration doesn't mean anything, since I have no idea what "claim" you're referring to. - Aoidh (talk) 20:22, 1 August 2014 (UTC)
The claim I was talking about is that of your edit message. Specifically: “That wording implies GNU uses it and that's it.”. I thought than it was clear in my message, but apparently it's not. I will elaborate to make it absolutely clear what is what I meant:
Your edit message can be interpreted in 2 different ways as I see it (Due to the ambiguity of “that's it”):
  1. The caption says that GNU (And nobody else) uses Bash. The caption don't implies that, it says “Bash, GNU's shell [...]”. Here GNU's means that GNU linguistically possess Hurd (As in “my country”), but that can't be correctly interpreted to mean that only GNU uses it. For instance, to say “Wikipedia, Wikimedia Foundation's encyclopedia” don't means that Wikimedia Foundation is the only one allowed to use Wikipedia; it can be interpreted to mean that Wikimedia Foundation is in charge of, and oversees Wikipedia, and that's the case between GNU and Hurd as well.
  2. The caption says that GNU uses Bash (And possibly others do as well), but don't says anything else about Bash. This isn't the case either, the caption that you reverted says “[...]widely used in Linux”. Hence it does say something else about bash.
Bear in mind that GNU isn't just the developer of Bash (If you want to call it so), it's the project that oversees it and many other software packages commonly used as parts of the operating system described in the article.
“but as explained, context is important and in the context of this article the caption you're trying to add isn't nearly relevant enough”. I'm not sure of what you mean by “as explained”, please clarify. I understand than it's your opinion that the aforesaid caption isn't relevant enough, but please provide a support for this claim.
If you have objections about my behavior I will consider them if you make constructive comments but let's try to keep this talk page on topic. For the record, regarding your comment “Yes, and telling people to "feel free to use the talk page" after they have already left a message there isn't helping.“ seems to me to be written in the form of an accusation rather than a constructive suggestion. I won't reply to it to avoid a discussion about the virtues and vices of either of us.
Regards, QrTTf7fH (talk) 22:19, 1 August 2014 (UTC).
Hello there! As Aoidh already described it above, saying "GNU's bash" actually decreases the value of bash and all the effort GNU has put into it. Also, if GNU isn't the developer of bash, what it is? Sure thing, it all depends on how one grasps the meaning of being a developer of something, but if something has been made under the GNU's umbrella, then GNU was obviously making (or, more specifically, developing) that thing.— Dsimic (talk | contribs) 04:47, 2 August 2014 (UTC)
  1. ^ Eckert, Jason W. (2012). Linux+ Guide to Linux Certification (Third ed.). Boston, Massachusetts: Cengage Learning. p. 33. ISBN 978-1111541538. Retrieved April 14, 2013. The shared commonality of the kernel is what defines Linux; the differing OSS applications that can interact with the common kernel are what differentiate Linux distributions.