>>>> "Ville" == Ville Skyttä
<ville.skytta(a)xemacs.org> writes:
Ville> I guess it can be done, but there are no guarantees that it
Ville> will stay in sync. For example, if we now bump idlwave's
Ville> package VERSION to 4.15, then notice a bug in let's say
Ville> Makefile or package-info.in, we'll need to release 4.16
Ville> though the idlwave inside is 4.15...
Right. In some cases, with very stable packages with an XEmacs person
running things, this is basically not a problem. For example, AFAIK
VM hasn't had anything but common versions for years now.
My packages (edict, mule-ucs, and latin-unity) don't really have
separate author versions, either (or at least didn't when Steve was
running things). We just made sure that the version bump got noted as
a maintainer update (presumably containing cool new or fixed
functionality) or an administrative update (Makefile/package-info
conformance, etc) in the ChangeLog. Basically when I wanted a
release, I'd bump the AUTHOR_VERSION to VERSION + 1, ChangeLog it, and
let Steve take care of the rest. Worked great for me!
I think that we should start moving toward a .deb/.rpm-style
AUTHOR_VERSION + XEMACS_VERSION numbering system (really, they are
different purposes and should ahve different numbers), but for the
moment as long as Carsten is willing to put up with _occasional_
non-consecutive author version numbering, there's no trouble even now.
--
Institute of Policy and Planning Sciences
http://turnbull.sk.tsukuba.ac.jp
University of Tsukuba Tennodai 1-1-1 Tsukuba 305-8573 JAPAN
My nostalgia for Icon makes me forget about any of the bad things. I don't
have much nostalgia for Perl, so its faults I remember. Scott Gilbert c.l.py