> 1. Development trees: each one will place the development tree under > his/her home: > ~<username>/scm/topic.git (e.g. ~mst/scm/sdp.git) > 2. Stable/release trees: these trees will be located under /pub: > /pub/scm/<name>.git (e.g. /pub/scm/ofed-1.1.git) > 3. All trees (development and stable) should be exposed via git web > interface. > > I think this methodology is similar to the way git trees are > handled in Linux, and its also a simple convention that will make > it easy to understand the type of each tree
If you mean the way git trees are handled on kernel.org, that's not true. There is only one class of git tree -- kernel.org happens to use /pub/scm but ~/scm/ is perfectly fine. But I don't see a reason to have two places to put git trees -- let's just pick one. I think trying to make a distinction between stable and development git trees is a mistake, since a single tree can contain many branches, some of which might be "stable" and some of which might be "development". In fact I think stability is a matter of degree so it may not even be possible to agree on whether a given branch is stable or not. - R. _______________________________________________ openib-general mailing list openib-general@openib.org http://openib.org/mailman/listinfo/openib-general To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general