My $0.02 regarding the whole 1.1 release naming:
I would have preferred that 1.1 not really be out there. I never
announced it and I couldn't remove it, but somehow it ended up as
the release for a bunch of systems. I never considered this a
"real" release myself.
The unfortunate fact is that although you did not "announce" the
release, by placing it in the public download directory it became a
release. It was a mistake, but not a big one. People would have used
whatever you placed there, be it an official release or a candidate
release. If you had called it RC4, RC3.1, whatever, it would still be
found in Debian Sarge, for example.
But as Jon has discovered, there are plenty of people out there who
will speak up when they think you're doing the wrong thing, but damn
few who will actually help. My advice to Jon is: do what you want,
and the rest of the world will have to suck it up if they don't like
it. You're doing the work: that gives you the right to make the
decision. I think he should give consideration to other people's
opinions, but the decision is ultimately his.
Jon made a slight oversight based the version naming, but again, not a
big deal. It'll get worked out one way or another. There was nothing
in his 09/02 post that indicated he didn't know about the existence of
nmh-1.1.tar.gz, just stating that, "This release should be the last in
the 1. line. After this, we should start a 2." I had understood that
to mean he would release a nmh-1.1.x.tar.gz tarball. nmh-1.2 or
nmh-1.1.x doesn't really matter. The confusion only exists by
re-releasing a 1.1 different than what's out there today.
Nmh-workers mailing list