Version 2.5.45 of the 2.5 branch of the linux kernel was released on October 30th 2002, two days ago. Highlights in this release include the already mentioned LVM2 volume manager and Digital Video Broadcasting layer [story], but also James Morris' cryptoAPI, the USAGI team's IPsec patches, Roman Zippel's new kernel configuration system [story] and the epoll syscall. Tons of minor updates and fixes have also went in since Linus resumed his work on the kernel after returning from the Geek Cruise [story].
In Linus' current bitkeeper tree, which will eventually become 2.5.46, another bunch of interesting features have been merged, including the Extended attributes / ACL for ext2/ext3 patch by Ted T'so, the hugetlbfs filesystem, the v4l2 API (a replacement for the aging Video for Linux API; for now, the original v4l still remains in the kernel tree, though), the Orlov block allocator for ext2 and an IPC update that, among other things, reduces lock contention, resulting in a 2 to 6% increase in the number of transactions per second measured by the OSDL's dbt1 test.
FreeSWAN
Does this mean FreeSWAN is dead or going to be dead?
Great work!
Simply Amazing.
You covered just about everything, schneelocke, from ACLs to epoll; and also stuff which i didn't knew had been merged (and some i didn't even know existed!)
More good stories like this one, please!
Re: Great work!
Thank you! I'm very glad indeed you liked the story, and I'll try to keep it up. :)
--
schnee
Re: Great work!
I second the sentiment - you're posting many good stories. Thanks!! I look forward to reading more... :) :) :)
Re: Great work!
Thank you! ^_^ It's really great to hear that people actually like what I write. BTW, I'm also very interested in becoming a mod; I just haven't gotten around to replying to your email yet.
--
schnee
Like always compilation fails
Like always compilation fails... :)
--I don't believe in fate...
I'm wondering how it is that
I'm wondering how it is that just about every 2.5 kernel I've downloaded has needed to be patched just to get it to compile. I realize that 2.5 is a development branch but shouldn't it at least compile?
NOTE: This is not a flame. Just a question from a beta tester and potential bug hunter.
Re: I'm wondering how it is that
One should think so, but in general, I tend to think of 2.5 releases as snapshots of current development rather than actual "releases" - similar to the nightly CVS tarballs provided by many projects.
But of course, you still do have a valid point.
--
schnee
Its because...
Those who are doing (non bk) development would much rather Linus get something out the door to sync with than have him spend more time trying to get everything to compile only to have it broken again. It should start getting better from now on.