On Wed, Nov 26, 2008 at 08:33:21 -0800,
Jesse Keating <jkeating(a)redhat.com> wrote:
My continued main concern with this though is fracturing the testing
pool, and not capturing enough eyes/minds on the frozen content for the
release. The sad fact of many releases is that many people seem to not
start looking at overall polish and bugfixing until the final freeze is
hit. They treat the final freeze as the last development day and expect
to have time after that to do bugfixing. Combining that with taking
users away from the QA on those frozen bits could lead to an even worse
release. So I think there are two problems to solve there.
I didn't find continual testing of the frozen bits all that useful. I wanted
to continue tracking F10 bugfixes especially xorg-x11-drv-ati and kernel
fixes since Dave Airlie was trying really hard to get KMS in shape for the
release and rapid feedback seemed help.
So I think there are three tracks that should be under consideration, not just
two. One is tracking the release, another release plus potential zero day
updates and the last the next rawhide.