On Jul 15 16:35, Karanbir Singh wrote: > On 07/15/2014 04:15 PM, Lalatendu Mohanty wrote: > > On 07/15/2014 04:37 PM, Karanbir Singh wrote: > >> On 07/15/2014 09:49 AM, Thomas Oulevey wrote: > >>> So my proposal is: > >>> > >>> For the build tags : <SIG><MAJOR>-<TAG>[-<COLLECTION>,]-build > >>> And the target : <SIG><MAJOR>-<TAG>[-<COLLECTION>,] > >>> The destination can be <SIG><MAJOR>-testing/<SIG><MAJOR>-release for > >>> all the target. (or keep the Fedora naming dist-*, but it looks more > >>> confusing to me) > >>> > >> I think this works, the only consideration is that we might need to ( we > >> really should ) give the SIG's a -devel a -testing and -production > >> target, so they can build and release into vairous repos. > >> > >> At the very least we would need a -devel where they can do whatever, and > >> a -testing where they can then push builds that need to be made public > >> on the repos, and once some criteria is satisfied, we can push content > >> from -testing to -production ( I dont think we need a rebuild for that, > >> just moving the rpms over should be good enough ) > >> > >> > > If we have facility to scratch build, do we need a "-devel" tag as we > > can do scratch build as many time we want? As I mentioned in a previous > > mail, we can promote a build from -test to -production after it meets > > certain criteria. Building it for multiple targets will be overhead IMO. > > that works too - the important thing is to have some process that people > can use to do these builds +1 for koji+bodhi. Rpkg already speaks both, but of course we can always do our own thing if we want. Possible, -1 for bodhi because I'm not sure the karma-gathering process maps well to the more centralized promotion workflow (presumably promotions fall to the core team?). Anyone with more bodhi experience care to chime in about karma? Brian -- Brian Stinson bstinson at ksu.edu | IRC: bstinson | Bitbucket/Twitter: bstinsonmhk