So was there a final conclusion? I get it that it makes sense to create a stabilisation branch shortly before really rolling a release (kind fo RC builds). This is basically a release branching attempt. But as we lack of a ReleaseManager who _reviews_ and pulls over all those changes on a daily basis I thin there is just no justification for this development branch thingy.
Any objections to merge development into master? LieGrue, strub > Am 31.01.2015 um 13:36 schrieb Thiago Veronezi <[email protected]>: > > Oh... I see. I'm not talking about feature implementation. I'm talking > about the release process and release branches. > I don't like a branch per feature either. :) We don't need that. > > > On Sat, Jan 31, 2015 at 7:31 AM, Mark Struberg <[email protected]> wrote: > >> If you push it to github or attach it as patch you will get help. >> That was also the way it used to be with SVN. And it would even be ok to >> grant an ASF colleague rights to your github repo or pull a fix from him. >> >> Of course it would be even better if we could have such branches in a >> 'sandbox' repo on ASF hardware. But we don't have those YET... :) >> >> >> LieGrue, >> strub >> >> >> >> >> >>> On Saturday, 31 January 2015, 13:28, Romain Manni-Bucau < >> [email protected]> wrote: >>>> No help yes. That is what we should target anyway. >>> >>> And removing a remote temp branch is not bad. >>> >>> Mark got a good proposal with staging mirrors of main asf repos. >>> >>> Le 31 janv. 2015 13:22, "Thiago Veronezi" <[email protected]> >>> a écrit : >>> >>>> @Romain: if we are talking about temp remote branches, it implies later >>>> removal of this branch [bad] otherwise we end up with dead branches >> [bad]. >>>> If we are talking about local or remote branches that can be used by >> only >>>> one person, it implies no help from the other developers [bad]. >>>> >>>> >>>> >>>> On Sat, Jan 31, 2015 at 4:46 AM, Romain Manni-Bucau >>> <[email protected] >>>>> >>>> wrote: >>>> >>>>> @Thiago: not exactly since we do it in temp branches which could even >>> be >>>>> private >>>>> Le 31 janv. 2015 10:32, "Mark Struberg" >>> <[email protected]> a écrit : >>>>> >>>>>>> The "Apache Way” is not a set of rules. The Apache Way >>> is >>>>>> >>>>>>> community over code. That’s it. Nothing else. >>>>>> >>>>>> >>>>>> Alan, it is important but all that comes only _after_ some very >>> basic >>>>>> legal rules we have to follow. >>>>>> Those are not many, but they exist. E.g. if a PMC doesn't >>> like to >>>> follow >>>>>> the license and marks guidelines then it will simply get shut >>> down by >>>>>> board. And I'm glad that we don't have such issues at >>> TomEE. >>>>>> >>>>>> >>>>>> To understand the importance of code provenance you just need to >>> look >>>> at >>>>>> the current subpoena we have to handle: >>>>>> >>>>>> >>>>> >>>> >>> >> https://blogs.apache.org/foundation/entry/the_apache_software_foundation_subpoenaed1 >>>>>> >>>>>> So it is really _very_ important for the foundation to have a >>> very good >>>>>> SCM history which can stand the proof of court! >>>>>> >>>>>> >>>>>> >>>>>> LieGrue, >>>>>> strub >>>>>> >>>>> >>>> >>> >>
