Talk:Linux (operating system)/Archive 3

How do we define success?

The opening line "Linux is the most successful open-source computer operating system in the world" is pretty subjective in my opinion. How do we define success? Number of people using it? Number of bugs? Maybe a line like "In terms of number of users and commercial support, Linux is the most successful open-source computer operating system in the world."
Just blurting out "Linux is the most successful" ignores other OSes who in some ways are more successful than Linux, depending on what one calls "successful." Eric M Gearhart

I agree. I was actually thinking similar thoughts when I read that last night. I think we need to re-write this article. (Check your e-mail for more info.) I'll try to re-write up to the point of what we have now today if I can. --Joshua David Williams 09:15, 19 April 2007 (CDT)

Concepts to be implemented

Please feel free to modify this list in any means necessary, and strikeout items that have been finished with the <s> tag. Also, please do not archive this section until all of these sections have been finished.

• The Cathedral and the Bazaar
• Trademark disputes
• Copyleft
• Database vendors switching to Linux
• Legality of codecs and free alternatives
• Devote an entire section to portability
• Lack of mainstream software support, especially gaming

--Joshua David Williams 21:50, 11 April 2007 (CDT)

Call for Approval

Per CZ:Approval_Process, as an author who has contributed significantly to this article (among others) I am requesting this article be submitted to the Approval process by an editor. --Eric M Gearhart 18:00, 11 April 2007 (CDT)

It sounds like there's still a lot of work underway. In my opinion, it's still too early. Greg Woodhouse 21:59, 11 April 2007 (CDT)

Although I think that what we have now is great, I too believe that there is plenty of information that could be added (see the checklist I added above). For now, I'm changing it back to status 2. Also, don't forget to leave the names of the previous people who've edited the checklist. Ciao! :) --Joshua David Williams 22:04, 11 April 2007 (CDT)

This article really is not ready to be considered for approval. While it contains very good information, the style of it is very problematic. Please see CZ:Introduction_to_CZ_for_Wikipedians#Get_ready_to_rethink_how_to_write_encyclopedia_articles.21 making sure to follow the links in the section. I tired to take a stab at it, but I am done for the night. Zzzz. Stephen Ewen 03:19, 12 April 2007 (CDT)

I would ask and encourage Computers editors to be more proactive with giving guidance such as this more often... half the reason I called for approval was to find out what this article still needs. --Eric M Gearhart 03:51, 12 April 2007 (CDT)

LOL, that's not to be encouraged! What actually brought me here was this forums post.

I also removed all the headers. One way to help create articles that are not modular is to remove them, and then write the whole narrative as a cohesive whole. The key headlines can be added once the article has reached or is near completion.

Stephen Ewen 05:04, 12 April 2007

I'll take a look at the post when the forum's back up... right now its returning "Sorry, SMF was unable to connect to the database. This may be caused by the server being busy. Please try again later." And yes I realize I was being evil with the call for approval :) We just need more support over at the Computers Workgroup --Eric M Gearhart 05:25, 12 April 2007 (CDT)

-sigh- Pat was right when he mentioned that we need to keep in mind the point to every sentence. We have lots of good information, but we're losing the point to everything we say. We need to come up with a structure for the article and re-write this. --Joshua David Williams 08:58, 12 April 2007 (CDT)

reorg of old sections, new sections needed

Here's some suggestions for where else to take this article. I reorged the old ones into two simpler sections, Origins and Applications. What I think would help, now, would be a section on Linux' successes and failures in the marketplace...I have now moved the long summary that used to be here into the article itself.Pat Palmer 22:16, 9 April 2007 (CDT)

Which is correct?

Is it Linux' or Linux's? Do any of you know which is technically correct?

As far as I know it would be Linux's, because although the name ends in an 'S' sound it's not actually an 'S'.
Compare with Alex' versus Alex's --Eric M Gearhart 15:03, 9 April 2007 (CDT)

Linux

Do you guys think the Linux controversy is notable enough to have its own article? --Joshua David Williams 22:36, 9 April 2007 (CDT)

I wouldn't break that out at this time (unless this article becomes too big). I think it's pretty well covered out on the larger world wide web. However, if it really turns you on, perhaps you should go for it.Pat Palmer 23:00, 9 April 2007 (CDT)

The history of the open source community is really my niche. I might have a go at it after I've finished the Tux article. Right now I'm having a bit of a hard time finding the correct chronological order for that one. --Joshua David Williams 23:02, 9 April 2007 (CDT)

The intro and history in the main Linux page make a good entry point for Linux related information. The Linux controversy should be an article of its own. Ari Hakkarainen

Kernel type

Shouldn't microkernel and monolithic kernel both redirect to Monolithic versus Microkernel? --Joshua David Williams 13:36, 11 April 2007 (CDT)

I'll take this silence as a "yes". If not, feel free to change the redirection pages. --Joshua David Williams 19:26, 11 April 2007 (CDT)

Done?

Are you sure we're ready for the #1 status? Compare the information we have to the Wikipedia article. --Joshua David Williams 19:49, 11 April 2007 (CDT)

I would leave it at 2 for now. I'm not done with this article.Pat Palmer 20:11, 11 April 2007 (CDT)

This is a grump

The "Origins" section now rambles and seems to lack purpose; it has become cluttered and uncertain. I originally organized it to center around the "upstart" Linus Torvalds taking on Tanenbaum. The important opening 2 or 3 sentences are--GONE, with no explanation. They've have been removed and a bunch of rambling stuff is now there. As a technical writer, I like to have a purpose in each section--a point to make. The point to make here, is that the creation of Linux by Torvalds astounded the computerati of the time. I don't want to start a revert war, but please think about whether it is "diffusing the purpose" when you go and just add facts to the top of a section someone just wrote.

I'd like to suggest that we start some kind of list, on this page, of random facts that we want to add, but aren't sure where to put them. I don't like to write about Linus Torvalds' intentions; I'm not a mind reader. I think the article should just talk about what happened, not what he was thinking. [Signature Missing]

I think making a list of facts is a great idea. But as for what he was thinking, I beg to differ. We have an ample supply of evidence to suggest the reasons he created Linux in the first place. I have not yet had a chance to cite the entire article, but this will be done eventually. --Joshua David Williams 21:33, 11 April 2007 (CDT)

Well the fact of the matter is that Torvalds didn't create Linux to "take on Tanenbaum." That argument between them happened later. I know it's not as dramatic, but Linux was started as a hobby, "nothing big and professional like GNU," that was picked up by the community. There's the dramatic "scoop" this section needs. --Eric M Gearhart 03:57, 12 April 2007 (CDT)

It seems like we have very different writing styles, and so I think we have to talk about writing styles, or else, I have to go away and work on something else for now. Please advise. Some of you know many more facts that I do. But facts listed willy nilly do not a story make. And Linux is a good story. Let's make it a good story, as well as factual.Pat Palmer 20:29, 11 April 2007 (CDT)

I do agree that the history of Linux makes a remarkable story, but please keep in mind that this is an encyclopedia, and not a novel. When you get a chance, please read CZ:Policy_Outline#Article_Standards. [Signature Missing]

Pat has a point though... our articles should "draw in the reader." See Life for a great example of an article that draws the reader in and keeps them "hooked." --Eric M Gearhart 03:57, 12 April 2007 (CDT)

"very successful as server, moderately successful as desktop; I'll bring statistics to prove this if asked"

No disagreement here, but you are asked for the sake of precision and citation. --Joshua David Williams 21:33, 11 April 2007 (CDT)

"Much older Unix operating system"

I think you need to be a bit careful here. Unix certainly did start out as a single operating system, but there is now no one Unix operating system. Some people prefer the phrase Unix-like operating system for just this reason, while others will refer to "flavors" of Unix. Still another point of view is that what does and does not qualify as Unix is a matter of genealogy of the code base, while others reject this point ofg view as outmoded or irrelevant. Finally, I think it's very important to resist the urge to make polemical statments about Linux vis a vis other operating systems. The very phrase "much older Unix operating system" is a bit of a slam, and not appropriate here. Greg Woodhouse 21:45, 11 April 2007 (CDT)

Yes this is a valid point - Unix is an OS in its own right. No need to slam it on the Linux article. --Eric M Gearhart 05:30, 12 April 2007 (CDT)

I worked at Bell Labs all during the '80's, into the 90's, and I worked with Unix flavors and taught it recently as faculty. I can personally testify that part of the attraction of Linux was the fact that its user experience is Unix-like. This is not an attraction I personally share; the command-line experience and shell scripting, of which I have done a lot, never did excite me, but I know that it does some people. The widespread desire fueling OSF through the '80's was for a Unix-like OS that ran on cheap hardware and didn't cost much. I have no problem at all saying "much older Unix operating system". As far as I am concerned, it should have died a long time ago but has been kept alive by people resistant to change. That said, I did grow to admire Linux and the energy and open source initiative fueling it, and I am now a fan, along with a lot of other slower "converts". There is no slam intended in saying that Unix is more than 30 years old, and Linux has a user experience like that. Think about it.Pat Palmer 21:40, 13 April 2007 (CDT)

Well, yes, and Unix has developed considerably since it's original introduction. A modern Unix is very different from the one first developed at Bell labs. Greg Woodhouse 08:16, 12 April 2007 (CDT)

Missing topics

• I see no mention of X11 and its role in graphical user interfaces for Linux.
• I see no mention of POSIX or POSIX compliance.
• I see that Wine is mentioned but not true virtualization (VMware, QEMU).
• I don't see any discussion of the Linux kernel and how it differs from traditional Unix kernels. What is innovative about Linux (kernel threads, etc.)?
• You do mention that Linux has enjoyed more success as a server platform but don't elaborate. What about Linux makes it a good fit here?
• You don't say anything about Linux on laptops.
• You really don't say enough about distributions. What makes Debian different from Ubuntu, Red Hat, SUSE, etc.
• You don't say much about groupware, office suites, etc.
• What are the advantages and disadvantages of Linux to a home user? a business user? a programmer? an artist?
• How might Linux fit into a heterogenous network?
• What about Linux security and SELinux?

Greg Woodhouse 08:39, 12 April 2007 (CDT)

This is where we need to take a step back and look at an outline of "Subtopics." In my opinion Linux can be a general description to start off with.

• I see that Wine is mentioned but not true virtualization (VMware, QEMU).

This would be in a Linux software or Software emulation article I'd think

• I don't see any discussion of the Linux kernel and how it differs from traditional Unix kernels. What is innovative about Linux (kernel threads, etc.)?

This would be in Linux kernel

• You don't say anything about Linux on laptops.

This would be in Linux probably

• You really don't say enough about distributions. What makes Debian different from Ubuntu, Red Hat, SUSE, etc.

This would be in Linux distribution

• You don't say much about groupware, office suites, etc.

Again this would be in Linux software

• What are the advantages and disadvantages of Linux to a home user? a business user? a programmer? an artist?

This would "fit in" the main Linux article

• How might Linux fit into a heterogenous network?

This would probably fit into a sub article like Linux and other operating systems, that would talk about Samba, Active Directory, and NFS etc.

Linus' intentions

I think only Linus Torvalds is qualified to talk about what his intentions were. If he has done so, please include a reference. If he has not done so, then those statements are rampant speculation.Pat Palmer 00:33, 15 April 2007 (CDT)

Can the linux kernel be called monolithic?

Can the linux kernel be called monolithic if it has modules? Aren't those modules something that is kernel-like, and is not necessarily part of the kernel? I mean, you can insert or remove them (well, not so much remove them in newer versions).

I don't know the terminology, it just seemed weird to me, so I wanted to ask. --Nick Johnson 15:39, 18 April 2007 (CDT)

Technically, it's called modular monolithic. The actual kernel itself is monolithic, and most of the drivers can be added as modules. Actually, most of the features can be compiled into the kernel itself, and aren't limited to modules. The ability to use modules is just a feature of the kernel, you might say. --Joshua David Williams 15:42, 18 April 2007 (CDT)

"Monolithic" in this sense refers to the debates over using a Mach-type kernel design versus using "one big binary." The Linux kernel is still monolithic, it just adds the "nice" feature of being able to dynamically load modules on the fly. The Linux kernel doesn't do any of the "abstraction into mini-servers that pass messages" that Mach, or Mach-like kernels such as Windows NT do. Eric M Gearhart

My compliments

This article is coming along nicely. I haven't looked in a few days; seems like it's really jelling. Keep up the good work!Pat Palmer 18:19, 19 April 2007 (CDT)