Latest Beta Release Over a Year Old?

[Stephen J. Turnbull]

Jake Colman writes:

 > I've been thinking about following the beta releases a bit more closely and
 > possibly getting more involved.  According to the web site, the last beta
 > release was 21.5.27 on May 16, 2006.  Has it really been over a year since a
 > beta release was done?

The honest answer is "yes."  It is true that 21.5.28 was released on
May 21, but I dropped the ball on posting the web message.

 > Also, is there a summary of the new features that will be included in the
 > next major release when the beta cycle is completed?

No.  A random selection of stuff that currently is in the mainline, in
more or less advanced state of progress:

1.  A major revision of Mule, with reasonably good support for Unicode
    at the code point level.

1A. We hope to switch to Unicode as the internal encoding, but no real
    work has been published on this yet.

2.  New garbage collection algorithm.

3.  Generational garbage collector to take advantage of the the GC
    framework.

4.  Piss-poor support for anti-aliased fonts via Xft on X platforms.
    It's beautiful and quite usable if you know your way around Xft
    and fontconfig, but from the point of view of the general user it
    hardly deserves more praise than "proof of concept".

5.  Canna (Japanese input method) support has been modularized.

6.  Some improvements in pixel-oriented window configuration.

6A. It's been suggested that we should do all the low-level stuff in
    pixels, and only provide character-based config as a user
    convenience.  No work done yet.

7.  Random fiddling with Mule stuff.

8.  Random syncs to GNU Emacs.

9.  A framework for "behaviors", which are an interface to global
    minor modes.

Stuff I'd like to see (and have done some work on, besides 1A):

10. Modules for curl, and neon, and maybe expat, libxml2, or gitlib
    support.  (libcurl and neon support are ready to go anytime, I
    think, except that I have no tests or application code for them.
    Nothing has been done for the others.)  Maybe modularize some of
    the other Japanese support stuff.

11. libsvg support for SVG images.  (Almost nothing done yet.)

12. Bootstrapping package installation from the core distribution.  (I
    just need to decide to put the time in.)

13. Whatever it takes to support AUCTeX and preview-latex properly.
    (Zero work on this yet).

14. Integration of Andrew Choi's Carbon port to Mac OS X.

Non-code efforts:

15. Move the repository to a decent distributed SCM, probably
    Mercurial.  This can be done any time we can find a host and
    people care enough to actually do it.  (I have a lot invested in
    git, and am unwilling to do the work myself until it looks like
    people will respond with a flow of patches.)

16. Set up an issue tracker.

The above is not a priority order.

Aidan Kehoe had some grand plans for Mule work, including 1A, but he
had the temerity to get a job.  Dunno what's up with the grand plans.
Ben Wing still posts once in a while and I know he has a pile of
pending code, but I don't know if he's really planning to ever do a
lot of work on XEmacs again.  Malcolm Purvis wants to get back into
GTK+ work, I think.

Blue sky stuff I'd like to see but can't honestly pretend I'll be able
to do: nuke font-lock and replace it with a real parser-based syntax
analyzer; do something sane about the extent mess (required for the
parser-based syntax highlighting); regexp engine overhaul; Unidata
support.

 > And is there any published timeline?

No.  I hope that doesn't discourage you from contributing.  There's
some hope that things will pick up in the near future, but nobody's
promising anything, just acting like they have itches to scratch. :-)