Isn't that what -current is for...

Pat doesn't usually provide kernel upgrades for stable releases unless is the case of security patches, which will still be provided. So why would it matter if we get the .37 or .38 branch?

My understanding is that, as narz says, we will still get security patches for .37, however no more bug-fixes for non-security issues.
New bug-fixes and features will go into .38 (or whatever the latest stable branch is).

.32 appears to be a special-case long-term-support branch. But distros like RedHat have teams of developers who back-port patches into the version of the kernel they are shipping. Not something I would expect Pat & co to be doing.

For some nvidia cards .38 works better with the nouveau driver. However, I don't mind compiling my own kernel, and the config files for .38 kernel in testing makes it a little easier for those who have never done it before. There is also other work a rounds for the .37 kernel which could be mention in changes and hints.

I find it unlikely that we'll get any real security patches unless they are critical and public since the patches would have to be manually prepared/tested with 2.6.37.x instead of getting upstream support. I don't advocate dropping to a long-term kernel at this point since dropping to .35 would lose a lot, but since there are no future long-term kernels in the pipeline right now it puts Pat in a bind. I think we may get uber-critical patches at the most. I don't blame Pat for this, and I certainly don't envy the decision he has to make here...it seems it's a lose-lose situation with no viable alternative.

2 members found this post helpful.

IF we exclude that 'totally uninformed' guy which 'unfortunately' is the stable kernels maintainer, believe that 2.6.38 series will be next LTS ...

0 members found this post helpful.