Am Dienstag, den 26.01.2016, 18:07 +0100 schrieb Emilio Pozuelo Monfort: > On 26/01/16 17:54, Gianfranco Costamagna wrote: > > Hi Emilio! > > > > > > > Can send another summary with your proposed plan here? Currently > > > libpng12-dev > > > provides libpng-dev but libpng16-dev doesn't. Are you going to > > > switch that? Or > > > do you plan to build a libpng-dev package from src:libpng1.6? > > > > > > sure, the only reason that libpng-dev wasn't provided has been to > > avoid people building accidentally > > with libpng16 when uploading on experimental. > > > > IIRC the plan is to upload on unstable with the Provides: libpng- > > dev line enabled > > (it is already there, just commented), > > and the change for the udeb file conflict (ongoing test right now). > > > > I think libpng12-dev will stop providing the libpng-dev at the same > > time, but I'm not sure > > Either that needs to happen, or a real libpng-dev package needs to be > built. > Otherwise if we have libpng12-dev and libpng16-dev providing libpng- > dev, things > won't be deterministic.
> > (My*) plan'd be to upload (1) libpng12 WITHOUT providing libpng-dev and (2) libpng16 WITH providing it. I'm not sure whats better: if that should happen at the very same time, or if we should delay (2) until e.g the -12 package arrived at all archs, to ensure that there will be no point of time where libpng-dev is provided by two packages or solve any races by delaying the start of the binnmus until we can ensure that all archs are getting the right one. A real libpng-dev package would be actually my* favorite, but this requires going through NEW and I fear that that would destroy the momentum we experience now to push this forward. Also, a real libpng- dev Package could still be introduced later. -- tobi * Nobuhiro / Anibal should give the definitive direction how the want to manage their package...
