* Hakan Tandogan <[email protected]> [2012-12-20 13:20:20]: > With (larger) code changes, I am absolutely of the same opinion as you > that there should be a review. The part I'm not so sure about is actually > mailing the complete diff to the mailing list. In my experience, creating > a fork and sending a pull request is a much more lightweight operation., > especially if you want to make sure that no email agent has messed up the > code change.
Actually git-send-email prevents that. > With the config files for Transifex, or even the translations themselves, > I would very much like to push them directly to the repo. I have to check > on the Transifex server, but I believe there were an option to only pull > reviewed translations. > > > Anyway, that might bring more agility to the developments. We will see. > > Maybe we could distribute the work on more shoulders. All for the best of > the project, of course :-) I think for translations, it's ok. For changes to the core (basically anything Python), there should be a review though. And as it was said, it doesn't go live, even on dev.m.o, without myself explicitely deploying the code there so we can always revert commits. Let's not put too much process in the way of contributions David, unless you can assure us you have time to dedicate to reviewing contributions quickly. I know I often don't... /Maxime -- Maxime Petazzoni <http://www.bulix.org> ``One by one, the penguins took away my sanity.'' Writing software in California
signature.asc
Description: Digital signature
