On Fri, 17 Sep 1999, Brian Mays wrote:
brian>
brian> > [EMAIL PROTECTED] (Brian Mays) writes:
brian>
brian> >> Once 2.2.12 makes it out of Incoming, we will have 8 kernel
brian> >> versions in the unstable distribution? Do we REALLY need to
brian> >> provide that many versions of the kernel??
brian>
brian>>>>>> "Guy" == Guy Maor <[EMAIL PROTECTED]> writes:
brian>
brian> > What about just keeping the last 2.0.x and the last 2.2.x ?
brian>
brian>That would be fine by me; however, some people might object because
brian>kernel "improvements" sometimes break things -- even in stable kernel
brian>branches. It is not so rare for someone to avoid upgrading to the next
brian>kernel version, because it breaks some obscure feature that he needs.
brian>
brian>Perhaps we should keep the last two versions of each branch? In this
brian>case, 2.0.35, 2.0.36, 2.2.10, and 2.2.12 (which is in Incoming). I
brian>don't know. Let's see whether anyone objects to just keeping two
brian>versions around.
In another thread, I am dealing with exactly this problem. My
machine hangs with 2.0.37 and 2.2.x, but is OK with 2.0.36. But had to
take a piece of driver code from 2.0.37. There are quite a few new
issues arising from two gcc branches and two stable kernel branches.
Having a few kernels around gives some flexibility in trying to put
together a working system. 11 kernels is probably too much, but a couple
of each might be OK. We (someone !) could also package the patches, which
is a bit more of a pain for the user, but we could get all 12 new kernels
without adding so much bulk to the archive.
John Lapeyre <[EMAIL PROTECTED]>
Tucson,AZ http://www.physics.arizona.edu/~lapeyre
