Linux in the news
All in one big page
See also: last week's Kernel page.
The current development kernel release is 2.3.16. This is another big patch, weighing in at over 4MB and touching over 1200 files. The bulk of the changes are in architecture-specific code and device drivers; there are some new RAID drivers in the system and Ted T'so's new serial driver has been integrated as well.
We have seen rumors that the 2.3 feature freeze will happen on Friday, September 3. Whether it hits on that exact day or not, it is clearly close. Then all that's left is the long slog toward a stable 2.4 release...
The current stable kernel release is 2.2.12, finally. Release notes for this version can be found on Alan Cox's web site. There is already a 2.2.13 pre patchout there with the next set of fixes to go in. It appears that 2.2.12 contains an unpleasant kernel memory leak bug that could accelerate the release of the next version as well.
Reiserfs is getting closer. Reiserfs is a longstanding project by Hans Reiser to produce a much more efficient filesystem for Linux, especially where small files are involved. Other nice features, like journaling, are also part of the plan. They claim better performance than ext2 - at least for some tasks.
Reiserfs came up this week when the development team stumbled across the (extensive) changes in the virtual file system layer in 2.3 - some things broke badly. They have been working through the difficulties, and expect to have something ready to integrate into 2.3 shortly.
For more information, a look at the Reiserfs web pages is worthwhile. There is a great deal of discussion of the motivation behind the project. And, of course, the ability to download the code.
A petition to get LVM into the kernel is being "circulated" at this web site. It currently boasts almost 200 signatures. The Logical Volume Manager, of course, is a management layer for disk subsystems which eases a lot of space management concerns.
An expression of opinion is, of course, a good thing. But Linus is just as likely, when presented with this sort of pressure, to dig in his heels as to include the code. The conditions for inclusion have always been a good implementation (as defined by Linus) and a clear need to have the code in the kernel. Attempting to replace those criteria with "special interest group" pressure is not the way to get a better kernel. Hopefully those wanting LVM in the kernel can discuss the matter with Linus and find out what, if anything, is blocking that inclusion.
Other updates and patches released this week include:
Section Editor: Jon Corbet
September 2, 1999