Cantor, Scott <canto...@osu.edu> writes: > On 4/29/19, 10:34 AM, "Boris Kolpackov" <bo...@codesynthesis.com> wrote: > > > The latter two are direct copies from the web/ and admin/ SVN directories. > > I believe that the web/ repository is actually directly published as > the web site, so there probably is additional work to do to change > how that's happening. I don't know the details of how it works, that > step to commit to svn was as far as I ever made it.
Hm, looking at admin/release-procedure.txt, step 15 in particular, suggest that it's at least not the whole process. In any case, I suggest that we convert it to Git now and decide what extra steps, if any, are required later. > > I think we should try to migrate the release tags (the ones > > in the Xerces-C_X_Y_Z form). I also wouldn't mind converting > > them to the now well-established vX.Y.Z form. > > I would cast a mild vote of preference for just using X.Y.Z for > the tag names rather than embellishing them at all with any other > characters, but otherwise +1 I have two reasons to prefer vX.Y.Z over just X.Y.Z: 1. This is the convention that is promoted (and sometimes required, like in Go) in modern languages and package manager. As a result, it is well recognized by tooling (e.g., GitHub will automatically recognize a tag like this as a release). 2. We may want to use X, X.Y, or X.Y.Z for release-series branches. We currently call such branches as xerces-X.Y but if you think about it, this 'xerces-' prefix is tautological (we are already in the xerces-c repository). So, to summarize, vX.Y.Z has potential benefits without any drawbacks that I can think of. So why not? --------------------------------------------------------------------- To unsubscribe, e-mail: c-dev-unsubscr...@xerces.apache.org For additional commands, e-mail: c-dev-h...@xerces.apache.org
