On Friday 20 October 2006 07:50, Christopher Blizzard wrote: > 1. Stop with FC6 and use that as a basis for development
I like this best of all. Far less chance of buildsystem screwage holding up OLPC builds / composes. Development trees have a tendency to break the build system from time to time (right Dave?) > 2. Follow FC7 development and expect that the timeline for FC7 will line > up with our shipping schedules (which isn't an unreasonable assumption > given that FC7 will be today + 6 months, I believe.) [1] Probably not the best of ideas. I know python2.5 is queued up for rawhide, plus a lot of other churn. Also, the schedule hasn't even been thought about yet, and if there is a possible merger of core+extras, that's probably going to have some effect on the schedule. > 3. Use FC6 + the RHEL5 kernel. Maximum kernel focus (hundreds of > engineers can't be wrong - right?), security updates are easy, things > don't move much. The problem you have here is if FC6 rebases the kernel and needs to rebase some of the userland to match. Then things may spiral out of being able to use the RHEL5 kernel. -- Jesse Keating Release Engineer: Fedora
pgpNHA6XEOIjP.pgp
Description: PGP signature
_______________________________________________ Devel mailing list [email protected] http://mailman.laptop.org/mailman/listinfo/devel
