Matthew Dillon wrote:
> Ok, I think we are homing in on it but there is too much confusion
> between stable and current... use ONLY CURRENT kernels. It doesn't
> matter if it's a debug kernel or not, but it should be a CURRENT
> kernel, not a STABLE kernel.
>
> -Matt
> Matthew Dillon
> <dillon@backplane.com>
Greetings:
Out of curiosity I tried the following I found on chlamydia:
2CSNAP-20041108-2330-GCC2-GENERIC-kernel.stripped.gz
2CSNAP-20041213-2330-GCC2-GENERIC-kernel.stripped.gz
Problem still present, so I see it has nothing to do with whether it is a
debug/non debug.
Concerning the stable vs current: point taken - my understanding (limited as
it may be) here is you are trying to track down the diff(s) where the
problem first appeared.
However, since the 2CSNAP only goes back to 20041001 I'm kinda stuck as far
as playing with kernels. My understanding here is that at some point in
time the dfly-stable-20041009.iso would have been -current before becoming
-stable. I am at a loss for knowing how to determine this point.
Since the dfly-stable-20041009.iso does not manifest the problem and the
dfly-20041107-stable.iso does, if the point where these were tagged stable
and packaged for iso distribution could be determined it would give a time
frame. Granted, it would be a month, but it would be starting from a point
when the problem was known to not exist - at least as far as my hardware
is concerned. I believe my box is a good test bed because I can easily
reproduce the problem within 6-10 seconds just by hammering the Numlock key
[or any other(s)] until loosing the keyboard.
Please forgive me for my lack of knowledge and sophistication regarding
software engineering. My intent is to try and help, but if all I'm doing is
adding to the signal-to-noise ratio then I'll just be quiet and listen. I'd
like to thank you for even responding to me at all, you just don't see this
kind of communication channel in a closed source environment and I'd like
to express my appreciation for its existence. It still kind of floors me
that one can actually communicate with a developer!
-Mike