>>>>> "vincent" == Vincent Danen <[EMAIL PROTECTED]> writes:
vincent> On Tue Jul 24, 2001 at 05:48:52PM -0700, Brian J. Murrell wrote:
>> > This question has been answered long ago and the answer is yes.
>>
>> I think we all know that answer. Yes starts to be belived as no when
>> it doesn't actually get followed up on with action however.
>>
>> How long has an 8.0 update kernel been "coming" for? It may sound
>> harsh, but IMHO, kernel building resources (people, server, whatever)
>> have no business working on "test" or "beta" kernels while there are
>> customers whose production systems are being impaired due to buggy
>> production kernels.
vincent> A while... unfortunately. The problem is that 2.4 is still really new
vincent> so fixes are being introduced all of the time, as are bugs. I think
vincent> the kernel team is trying to find a happy medium where the kernel is
vincent> stable and works without newly introduced bugs. The best way to test
vincent> this is via cooker, which is why cooker kernel development is so
vincent> fast-paced.
Now waiting for supermount to be stable, as we need it for an update,
it hasn't worked too well lately, but now it don't work at all :(
Then we got the ECN problems, drm problems, and the spec file is still
way too complicated (it is more cleaner know, but still not anything
that a normal person wants to read :(
vincent> I agree, to some extent. The best/test kernels are needed to find a
vincent> good kernel to supply for 8.0 updates. Again.. because 2.4 is still
vincent> relatively new, bugs are introduced and fixed with each new revision.
vincent> We have to find the best kernel to update with. It does no one any
vincent> good if we release a kernel in updates to just have to do it again
vincent> right away due to another bug.
I agree here with both of you. 2.4 kernel is a bit buggy, and it is
difficult to find a version that works well for everybody :(( We got
the VIA problems, we need to update kernel for that, same for adaptec
(aix7xxx problems), then we have the iptables problems (no way to get
an iptables package that works with linus kernel & iptables patched
kernel, and that is not extensively documented elsewere.
Other problem is that we don't always have _all_ the machines for
doing the testing, that means that a lot of times all that I can do is
upgrade, downgrade the driver, as normally the problems are hardware
bugs :(((
Same happened with tulip & 3128too, the people whose cards stooped
working are complaining but there are also a lot of other people for
which now their cards work.
>> > and I cannot answer that since I am
>> > not in charge of the kernel. But there will definately be a kernel
>> > update for 8.0.
>>
>> WHEN?
Pretty soon (tm).
Once that supermount works. I expect it for the end of next week to
have a kernel candidate.
>> I really really really DO NOT want to see another Cooker kernel until
>> there is an updated production kernel. Sounds like the kernel "guy"'s
>> manager needs to set some priorities and objectives.
I can release a kernel just now if that is what you want, but you want
it also working right?
vincent> Fair enough, but many people do care about this (myself included).
vincent> I'll find out what I can from the kernel team and see if we can't get
vincent> something into updates as soon as possible.
Sorry for the inconveniences & delays in getting a new kernel, but
there have appeared a lot of bugs in the kernel, and when I think that
I have a kernel ready for the update somebody appears with new bugs
that invalidate the previous work :(
Later, Juan.
--
In theory, practice and theory are the same, but in practice they
are different -- Larry McVoy
