On Mon, Mar 14, 2016 at 11:29 AM, Kamil Paral <kparal@xxxxxxxxxx> wrote: >> releng flipped any necessary switches for >> 'release' behaviour. > > I think this is the only meaningful argument for having TC/RC distinction. Personally I'm fine with it both ways. It's nice to see which images are the ones "almost ready", but I'm not sure it's worth having more bits for releng to remember and flip. > It was my understanding that with full nightly composes TCs were going away because basically every nightly compose was the equivalent of a TC. > If we keep the TC/RC separation, I like the approach above. > >> >> I think this still works a bit better if we add a 'Final' milestone >> rather than using the 'RC' milestone, but yeah, this isn't bad. > > "Final" would be nice. We can't do that without recomposing everything once it's been signed off. Basically any RC has the potential of becoming a GA release, we can't just rename files once we decide something has been signed off due to filesystem labels, checksums and all of the verification bits. So we would then have to recompose and re QA and hope the wind is still blowing in the same direction and it all comes out the same again, and if it doesn't we have to do it again and still call it "Final". So what ever the RC naming ends up being it has to be what ever we'd use for the final naming. Peter -- test mailing list test@xxxxxxxxxxxxxxxxxxxxxxx To unsubscribe: http://lists.fedoraproject.org/admin/lists/test@xxxxxxxxxxxxxxxxxxxxxxx
