On 6/21/13 10:45 AM, luc wrote: > Le 2013-06-21 17:34, Ralph Goers a écrit : >> I prefer the distinct separation between sandbox and proper >> (although >> I've never understood why the word "proper" was chosen). > > I agree for two different reasons. > The first one was already given by Sebb: users expect commons > proper to be mature and maintained. > The second one is that we have many components, so the separation > in three groups related to maturity > helps when someone looks at the global list for the first time or > only from time to time. With tens of > components, having to go to each one and identify whether it is > experimental or not is tedious. Our > current settings allows understanding the status at a glance.
+1 Regarding the process, [1] is a little out of date, but IMO the spirit is correct. We used to require PROPOSALS for all commons proper components and no longer do that. We also used to require that 3 commons committers express interest / willingness to develop a graduating component. We no longer require that. What we do still require is that the sandbox component be close to ready to make a release and there is a successful PMC VOTE for promotion. My personal opinion is that we should move a little back in the direction of trying to ensure we have viable communities around the components that we promote into commons proper. I am fine with making it easy for ASF committers (or any other volunteers) to get karma; but "committed committers" is what we need to grow to support new components - however they come to us - and unless and until we see a proto-community of such volunteers available, we should not promote things out of the sandbox. Phil [1] http://wiki.apache.org/commons/CommonsEtiquette > > Luc > >> >> As for moving components I would suggest starting a DISCUSS thread >> and then, if appropriate, move to a vote thread. >> >> Ralph >> >> On Jun 21, 2013, at 7:22 AM, Emmanuel Bourg wrote: >> >>> Le 21/06/2013 16:02, sebb a écrit : >>> >>>> Also, I don't think we should have code under proper that is not >>>> likely to be supported going forward. >>>> >>>> Users have the expectation that code under proper is mature and >>>> supported. >>> >>> It's doesn't prevent announcing a component as experimental with >>> a clear >>> distinction on the website. My proposal is simply to drop the >>> technical >>> distinction between proper and sandbox: same base directory in >>> Subversion, same ACLs. A component would simply "graduate" by >>> reaching >>> the version 1.0. >>> >>> Emmanuel Bourg >>> >>> >>> >>> --------------------------------------------------------------------- >>> >>> To unsubscribe, e-mail: dev-unsubscr...@commons.apache.org >>> For additional commands, e-mail: dev-h...@commons.apache.org >>> >> >> >> --------------------------------------------------------------------- >> >> To unsubscribe, e-mail: dev-unsubscr...@commons.apache.org >> For additional commands, e-mail: dev-h...@commons.apache.org > > > --------------------------------------------------------------------- > To unsubscribe, e-mail: dev-unsubscr...@commons.apache.org > For additional commands, e-mail: dev-h...@commons.apache.org > > --------------------------------------------------------------------- To unsubscribe, e-mail: dev-unsubscr...@commons.apache.org For additional commands, e-mail: dev-h...@commons.apache.org
