I was about to link it when I read the posts above

btw his Computer Networks was fine.
in that discussion I think he was just professor-bulling but he didn't know what he was dealing with
I'm just happy to be here and LOL at him

Now that was an interesting read form the past!

Sure, then don't reboot your machine :P I just feel that just because someone can maintain a stable branch longer, up to 2.6.35.101 for minor revisions, doesn't mean it actually was worth it to call it "stable" and put out so many revisions because people aren't going to install them and reboot that frequently. You have mentioned the same that you do not intend to make these as regular updates either. So what is called stable for a production server is arbitrary, yet I do hope that the best is picked for everyone in a distribution. Yes I will never prove that living on the edge for my desktop is stable either--I just want to finally test r600g which I don't think 2.6.38-rc2 is golden.

My feeling is this LTS junk that is being spoken of is garbage, especially in ubuntu. From what I remember they stabilized on 2.6.32 last year, yet then they turned around and started wholesale backporting from 2.6.33. They should have just given everyone 2.6.33 at that point, yet 2.6.32 was still called an LTS, and that's what they convinced people they have. Me, with my server hat on would have ran, far away, to slackware.

At least in the case of 2.6.35 they were right, 2.6.36.x was causing freezes on my system and I couldn't even run 2.6.37 long enough to test it because my screen goes blank. When major new features are introduced it is reasonable that the earlier versions containing them have more problems. In the 2.4 labeling scheme these intermediate versions would have been the development versions. I like the fact that 2.6 is released more frequently but 2.4 release versions used to be more stable.

2 members found this post helpful.

What you also have to remember about these long-term kernels is that they're a best-efforts backport of fixes from the current kernel. This is never going to provide the same degree of stability/security as officially maintaining two or more releases at the same time (FreeBSD style), but for those who simply can't risk following the bleeding-edge it is better than having no updates at all.

Of course, maintaining multiple releases means more work for the developers so I can understand why Linus doesn't want to do that. He's a busy man as it is.

2 members found this post helpful.