Re: Emacs 21.4

[David Kastrup]

Richard Stallman <address@hidden> writes:

> I have made an Emacs 21.4 release with a single security fix.

We noticed.

> So our coming release won't be 21.4.
>
> I now think it should be version 22.
> It has plenty of new features.

With regard to architectural features, we basically have GTK support
under X, image/toolbar/whatever support under MacOS and Windows,
Unicode in menus (though probably not consistent in all ports),
definitely improved images (functionality/performance).  Last time we
were thinking about version numbers, I was in favor for calling the
new branch release 22, partly to avoid a situation where important
fixes mandated an interim release, partly to bring across the point
that 21.4 would be nothing like 21.3.

It is unfortunate that there has been _lots_ of talk in the mean time
about what 21.4 will be, and references with "will work for Emacs >=
21.4" are on the web and software packages and manuals in hundredfold
as can easily be verified by Google.  So we'd better avoid this sort
of thing in future if we can.

What needs to be done now?

a) 21.4 has to be announced on all relevant channels and people told
   that it contains a single security fix and not what people thought
   21.4 would contain.  This is also important because the
   documentation of Emacs-21.4 is not in any manner updated: people
   will think they were the victim of a hoax upon
   downloading/installing an Emacs-21.4 distribution if 21.4 does not
   get announced publicly soon.  The distribution itself does not tell
   this too visibly.  It may be a good idea to announce that the next
   _major_ release will be 22.1 after all.

b) In the CVS version, the version number should likely get bumped to
   22.0.50 instead of the current 21.3.50.  There is no sense in
   making it 21.4.50, anyhow, and it would be quite misleading to have
   it stay at 21.3.50, lower than the current released version.

c) all references to 21.4 need to be replaced by referring to 22.1 in
   our CVS.

d) anybody speculating about version numbers for Emacsen having
   internal Unicode representation or multitty support must be gently
   but firmly locked into the freezer.

e) get drunk.

Personally, I volunteer to tackle task e) right now.  With a
vengeance.  Anybody willing to tackle task d) should wait until I
release the freezer.

-- 
David Kastrup, Kriemhildstr. 15, 44793 Bochum