Story: Is Linux about to fork?Total Replies: 10
Author Content

Nov 19, 2004
10:09 AM EDT
The article is confusing. It mentions that Linus may apply these large patches to the kernel for 2.7, but doesn't explain how that creates a fork. If 2.7 never becomes the stable 2.8, and instead becomes something else, like Linux Server, then maybe it is a fork, but the article never says that. Instead, it goes off on a tangent about Solaris. I don't get it.

Nov 19, 2004
11:43 AM EDT
Sun propaganda? ;-) Didn't read it, will not read it - repeats the same bogus worries. Oh, wait! Are they talking about Microsoft Linux? Then it might be possible. MS against the Linux World. Wait don't they already advertise in the rag ... mean mag. This is getting confusing.

So reread it - I am sure there is some hidden wisdom there!!

Send me a summary when you finish.

Nov 19, 2004
11:43 AM EDT
Here a quote: "Each version of the kernel requires applications to be compiled specifically for it."

Well, if this was true I had to recompile all my application software every time I upgrade my kernel. That boy either does don't know what he is talking about, or he is deliberately getting things wrong. The way that stupid article is written I suspect the latter, it is just there to confuse people and make them afraid of GNU/Linux.

Nov 19, 2004
11:46 AM EDT
Maybe the author doesn't understand how linux is developed. Or maybe the author doesn't want to. But regardless, a development branch is not a fork.

People have predicted forks for years. They have been wrong for years. The best features will survive, the worst features will die. There is no incentive to have two competing good implementations, unless there is a major breakdown in philosophies between two or more major groups, who go their own way. Either group can easily (and leaglly) add any features the other group produces, so that would be a waste of effort. No major philosophical breakdown appears to have happened, and the new features mentioned in the article are not dividing the community. They just don't want too many major changes to a stable codebase and so work for the future version on a possible branch.

Nothing really to see here.

Nov 19, 2004
12:03 PM EDT
There are tons of forks of linux already. They just happen to be binary compatible. What differs are the libraries that are installed.


ARM, PowerPC, x86, Mips, Each of these is a fork, at least according to this guy.

Also I don't know about any one else, but you can get any x86 binary to run on any x86 system easily. The hard part is linking to the libraries and making sure the proper ones are installed

Nov 19, 2004
3:05 PM EDT
Ah, but ARM, PPC, etc are all contained within the one source tree. This guy seems to be saying that linux will "fork" into a stable (2.6) tree and a testing (2.7) tree. I fail to see how that's a problem, though. They've been doing that for years.

Nov 19, 2004
3:46 PM EDT
The inside rear cover of the Linux Journal in November 2004 had a good article explaining the new development process and why there may not be a 2.7 kernel anytime soon, if ever.

Things have changed and not everybody has caught on yet and some like the author of the Linux fork article are confused.

Nov 19, 2004
5:28 PM EDT
Here is a write up on what's going on with the Development process since the Linux Journal issue is only available in print so far

Nov 20, 2004
2:43 PM EDT
I was going to make a wonderful original post but chris beat me to it.

"People have predicted forks for years. They have been wrong for years."

That sums it up pretty well.

Nov 20, 2004
10:39 PM EDT
This story doesn't make any sense. The author begins with ominous talk about Linux forking, but the quotes from Andrew Morton are clearly talking about starting a 2.7 *branch*. I think this author is just clueless.

Nov 21, 2004
3:22 PM EDT
It's been discussed on groklaw:

Posting in this forum is limited to members of the group: [ForumMods, SITEADMINS, MEMBERS.]

Becoming a member of LXer is easy and free. Join Us!