On Mon, Dec 12, 2016 at 6:04 PM, Jason Ekstrand <ja...@jlekstrand.net> wrote: > On Mon, Dec 12, 2016 at 8:57 AM, Marek Olšák <mar...@gmail.com> wrote: >> >> On Mon, Dec 12, 2016 at 4:46 PM, Nicolai Hähnle <nhaeh...@gmail.com> >> wrote: >> > On 12.12.2016 16:41, Daniel Stone wrote: >> >> >> >> On 12 December 2016 at 15:28, Emil Velikov <emil.l.veli...@gmail.com> >> >> wrote: >> >>> >> >>> As mentioned by others - having the second number represent the month >> >>> would be better, afaict. >> >>> Namely: YY.MM.PP. Thus 17.02.01 provides direct and clear feedback >> >>> that >> >>> - 2017 release, from the second month (Feb). >> >>> - first bugfix release. >> >> >> >> >> >> Not being funny, but does this mean that 17.02 bugfix releases would >> >> have to all be done in February, or could yyyy.mm.xx with xx > 0, mean >> >> that the release was not done in that month, but just the branching >> >> was? >> > >> > >> > While I think the answer to that _should_ be obvious (just look at >> > Ubuntu >> > LTS version numbers...), it is one reason why I'm not too keen about >> > using >> > the month. I'd say YY.AA.PP with YY = year, AA, PP = simply incrementing >> > should be good enough and avoids silly questions like do we take the >> > month >> > of the release or of the first -rc? What if the release slips into the >> > next >> > month by a few days? >> >> I second that. YY.AA where YY=year, AA \in {0,1,2,3}. > > > Agreed. It also reduces confusion about what to expect the next version > number to be and "where did 17.3 go? Oh wait, there wasn't one". Ubuntu is > consistent enough that they always do .4 and .10 so people know exactly > what's coming. We aren't that consistent so I think having it be YY.MM for > some random set of MM will just confuse people. > > Marek, Did you mean YY.AA or YY.A? I don't think we need two digits for the > minor number.
YY.A Marek _______________________________________________ mesa-dev mailing list mesa-dev@lists.freedesktop.org https://lists.freedesktop.org/mailman/listinfo/mesa-dev