I think new builds is a bad idea, a response to a worst-case event that hopefully never happens. If the quality assurance process that generated multiple alpha, beta, and release candidate builds has failed, another try to fix one more bug will have less complete testing: it is too likely to add new problems while it fixes others. Instead, users unable to install a new release should continue with an older Fedora release. It will only be six months until the next regular release. The truly adventurous (desperate) may experiment with Rawhide. For problems that impact more than very rare installations, someone is likely to create an ad hoc Fedora build for that group. This may be helpful, but it should not be described or distributed as a normal Fedora release. If problems with a new release are unusually severe, some of the effort that normally would focus solely on that release (upstream maintenance, new applications, kernels) may spread into earlier Fedora releases to help affected users wait for the next regular Fedora cycle. Only if these alternatives appear inadequate to a group of testers and developers obviously large enough to insure production-quality tests of a post-final new build should we attempt such an extraordinary measure. -- test mailing list test@xxxxxxxxxxxxxxxxxxxxxxx To unsubscribe: https://admin.fedoraproject.org/mailman/listinfo/test
