Linus took almost three weeks to get 2.6.24-rc4 out - the 2.6.24 process has, in general, been quite slow-moving this way. Normal practice is to try to get -rc releases out once per week. It only took me two days to update the corresponding forecast page, so I feel like I’m doing pretty well. The Btrfs entry has also seen minor updates - version 0.9 just came out. Things are getting better there, Btrfs almost doesn’t crash when the disk gets full…
Time for a few basic updates, starting with a somewhat belated acknowledgment that 2.6.24-rc2 is out. Now that I’ve caught up, expect Linus to release -rc3 just about any time.
PID namespaces continue to present complicated issues - it’s hard to present a coherent view of the system while simultaneously putting up walls between groups of processes. The result is that the PID namespace code may well not be available in 2.6.24, even though it will be present in the tree. I’ve added pointers to a couple of articles explaining why.
Peter Galli printed a fair article questioning Microsoft’s slicing and dicing of raw IDC numbers.
Here is the full text of my response to his original article:
Fundamentally this particular study will over-count Windows share and undercount Linux. Al Gillen at IDC, who we have a lot of respect for, says this himself in your article. Why is Linux so under-counted in this research?
The first 2.6.24 prepatch is out, signaling the closing of the merge window. So I have gone and tweaked things all over the Weather Forecast pages to match the new reality. The summary on the main page now shows the highlights of 2.6.24, and various other entries have been changed to reflect what happened during the merge window. In particular I quietly changed the predicted merge dates for all the stuff which I thought would get into this kernel, but didn’t.
We’re talked quite a bit in the press about how Linux is the platform of choice for .com and Web 2.0 development. Google, Amazon, Facebook and so many of the Internet’s leading applications or services are built on Linux and open source. And why is that?
Matthew Mengerinkm VP of core technologies at Pay Pal lays out some very clear reasons why Linux is the defacto choice in this article on Linux Insider.
The x86 merger has happened, so the miscellaneous topics page has been updated accordingly. While I was at it, I added an entry on KGDB - another attempt to get interactive debugger support into the mainline kernel.
I’m a little behind in my posts, but I wanted to throw my hat in the ring of cheer for the news that Novell is making Greg KH a fellow and allowing him to work full time on the Linux Driver Project. This really is excellent community support by Novell.
The ever useful Kernel Trap has the story here
Internet news has their take (and mine) here.
SMACK is a simplified mandatory access control module for Linux. It seems likely at this point that SMACK will have the honor of being the first such module to join SELinux in the mainline kernel. So I added an entry on the security page about it.