Re: 2.0.31 : please!

Alan Cox (
Sun, 13 Jul 1997 23:02:23 +0100 (BST)

> >Not unless you want to reboot every day, not run certain cards and
> >disable identd.
> I run this patch, I don't reboot every day, and I run certain cards.
> Am I doing something wrong? :-)

Yes. You aren't a valid sized test sample.

> Seriously - it appears that a critical mass of developers has decided
> that a stable, working kernel isn't worth the bother. I find this

You mean it appears that a critical mass of users don't appear to be
wishing to contribute to the code. A lot of the 2.0.3x stuff has been
very difficult to resolve, and 2.0.31 has to be very stable. 2.0.30 should
IMHO have never occured, it was forced on people by the perpetual whining
of some folks for 2.1.x features in 2.0.x.

The memory stuff is now basically resolved, there are a couple of inode
issues to sort, the kernel network hash stuff has a bug or two (hence
the identd Oops logs people sometimes see) and the drivers are also
awkward. The new tulip works better for a lot of boards but not all for example