2012/12/12 Andreas Tille <[email protected]>: > On Wed, Dec 12, 2012 at 09:12:11AM +0100, Rafael Laboissiere wrote: >> >Uploading to experimental in freeze time is handled in Debian Med >> >team the same way but we did not used a different branch. I'd >> >consider this interesting however, you might blame my Git beginner >> >status responsible for not understanding how you upgrade from >> >experimental to unstable once stable is released. Will you merge >> >the experimental branch with master then or will you simply move >> >on in master branch (targeting at unstable) by ignoring the >> >changelog entries done meanwhile in experimental. I would not >> >like the latter because this would be a loss of information for >> >non-Git users just looking at the package changelog. >> >> I think that merging experimental and master is the appropriate way >> of handling it, since the entries of both branches will appear in >> debian/changelog. > > OK. For those who care about the Git workflow in our team policy > document[1]: Would you consider adding the way the Debian Octave team > is handling the experimental branch to this document. (And if you agree > in principle but have time constraints please ping me.)
This is a sane workflow which allows continuing to work on the Experimental branch, and leaves the option to fix RC-critical bugs in Testing without taking all the experimental changes. I approve it. Although I think this makes only sense for packages already in Testing. A completely new package should just be done in the master branch. +Emilien -- To UNSUBSCRIBE, email to [email protected] with a subject of "unsubscribe". Trouble? Contact [email protected] Archive: http://lists.debian.org/canqxmqg8gvqczvwnafuhncxr5impps8hd4mdnfn2v5oyzhw...@mail.gmail.com
