Ok, what's up with the Alan Cox kernel versions?

[lsibn]

Sorry to have to ask this, but I've wondered for some time, and never actually seen anybody come out and say "This is what it is."

I think Alan Cox maintains a slightly modified Linux kernel, and I'm thinking this is what it is, but is that right?

Is there any special reason the ac releases are there at all, when they could be merged? Is it just that the ac releases are used as 'point-point-point' releases? (I.E., 2.4.8-ac2 = 2.4.8.2)

It's just something I'm very curious about, thx :D

[posterboy]

I believe that AC is a kernel manager, who codes and manages with the blessings of those who write for the kernel, much as Linus did himself long ago. AC gets say-so over what patches actually go into the kernel, and the AC releases are just that, a way to get the changes out so others can play with them, prior to final decisions about what stays in and what doesn't. This action in no way represents a fork, rather it is the process toward the released kernels. In the end, Linus blesses the work of AC, who has blessed the work of the other contributors, and we get a new kernel. HTH, Ray

[lsibn]

I didn't mean to say that the ac kernels are forked (because given the licensing and everything else, you'd be crazy to NOT merge the changes into the regular version).

Thanks for your explanation, though. Something I've been very curious about recently...

[stiles]

The AC series is a fork (well according to Linus, it's not a bad word or anything), and so is the AA series, they are separately maintained kernels (some patches will be accepted by one and not the other, this happens often). Yes there are constant merges from the Linus and AC branches, but there are time when they don't see eye to eye at all (do some search on major and minor device numbers, and what some hackers think of Linus' position and AC's). But it's all good, it's just how it's done. When Linus moves on to the developement branch, AC will be the stable branch maintainer.