[LWN Logo]
[LWN.net]

Sections:
 Main page
 Security
 Kernel
 Distributions
 Development
 Commerce
 Linux in the news
 Announcements
 Back page
All in one big page

See also: last week's Kernel page.

Kernel development


The current development kernel release is 2.4.0-test10. We could report on Linus's comments from his announcement of this release, but they really stand on their own:

Ok, test10-final is out there now. This has no _known_ bugs that I consider show-stoppers, for what it's worth.

And when I don't know of a bug, it doesn't exist. Let us rejoice. In traditional kernel naming tradition, this kernel hereby gets anointed as one of the "greased weasel" kernel series, one of the final steps in a stable release.

There has been the occasional objection that 2.4.0-test10 is not ready, but those have been very small in number. It looks like we're getting close.

The current stable kernel release is 2.2.17. The 2.2.18 prepatch is up to 2.2.18pre18. This one comes with more than the usual number of warnings, due to the deep nature of one of the fixes; if anybody has been burned, however, they are keeping very quiet about it.

Timpanogas Research Group appoints Andre Hedrick as CTO. The Timpanogas Research Group has announced the hiring of Andre Hedrick as CTO. Andre is best known as "the Linux IDE guy," the person in charge of the IDE disk subsystem.

LynuxWorks patents loadable modules? LynuxWorks recently filed for an IPO. The casual reader of the (2.5MB) IPO filing might easily have missed this little bit of interesting material:

We have developed a patented technology that enables LynxOS to be configured to leave out portions of the operating system not required for a given application, thereby reducing the amount of memory used. [...]

In addition, we have filed a patent that covers technology that allows the developer to extend the functionality of the LynxOS kernel without modifying the source code by adding modules for new functionality.

We have not yet succeeded in either finding the relevant patents in the database or getting patent numbers out of LynuxWorks. But from this description, it sure looks like the company is talking about loadable modules. Such modules, of course, are an important part of most Linux installations, and have been a feature of many operating systems for a very long time. It would be surprising if LynuxWorks had really come up with something new in this area.

As always, though, these patents are worrisome. It is getting harder to build a system without encountering more and more of them.

A new Linux IPv6 project. The USAGI Project announced its existence this week. USAGI stands for "UniverSAl playGround for Ipv6"; the project is digging into the Linux IPv6 stack with the intent of bringing it up to modern standards. That work includes bringing the code into compliance with a number of IPv6 RFC's, integrating IPSEC, and a bunch of other stuff.

The project's first release was also part of the announcement. It's based on 2.4.0-test9, and includes a number of enhancements and bug fixes to the standard IPv6 implementation. They have even made an IPv6 version of khttpd. Further releases are planned on a two-week schedule.

More information may be found on the linux-ipv6 web page. (Thanks to Per Harald Myrvang).

/proc in XML? A posting from somebody identified as "Joe" raised an intriguing idea: why not have the /proc filesystem return data in XML format? In this scheme, a file like /proc/meminfo would have lines like:

	<MemTotal>63036 KB</MemTotal>
Instead of the current:
	MemTotal:     63036 kB
Such a scheme would have some advantages. Almost every language now has a library that makes parsing of XML easy, so parsing the /proc entries should not be a problem. If the new format were properly done, the past problems where changes in the format of /proc files breaks applications should happen no more. And, most importantly, think of the great buzzword compliance that an XML-generating kernel would provide.

On the downside, this would be yet another /proc format change that breaks applications. And all that XML boilerplate would bloat the kernel image a bit.

But the real reason why this change will never happen is that the vision for /proc is a bit different. The plan is to split apart /proc entries so that each contains a single value, replacing file formatting with a directory structure. There's no need or place for XML in such a scheme, and parsing problems mostly just go away. But it's a fun idea...

Other patches and updates released this week include:

  • Gary Lawrence Murphy posted an updated call for participation in his KernelWiki collaborative documentation project. KernelWiki is coming along, a fair number of people have contributed material.

  • SUBTERFUGUE 0.1.99 ("a foundation for building tools to do tracing, sandboxing, and many other things") was released by Mike Coleman; it was followed quickly by a bug-fix update.

  • H. Peter Anvin announced a new policy for cryptographic software on kernel.org, in response to a further liberalization of U.S. crypto export policy.

  • Jens Axboe released version 0.0.2d of his packet CD writing driver.

  • A new user-mode Linux release was announced by Jeff Dike.

  • Bartlomiej Zolnierkiewicz has released a backport of the 2.4 IDE patch.

  • Andreas Gruenbacher posted a new version of his extended attributes proposal.

Section Editor: Jonathan Corbet


November 2, 2000

For other kernel news, see:

Other resources:

 

Next: Distributions

 
Eklektix, Inc. Linux powered! Copyright © 2000 Eklektix, Inc., all rights reserved
Linux ® is a registered trademark of Linus Torvalds