I've updated my code to the latest Emacs sources, and changed the
build process so that it uses an installed version of Guile instead of
a private, hacked copy. I've also got a public repository available.
My testing is mainly x11 builds using guile 1.8.7 on Mac OS X; no
Windows, no NS, but I'm planning to try building a Guile version of
Emacs.app soon; also with Guile 1.9.x (which still doesn't build
cleanly for me yet) and the BDW-GC branch.
I've got a public repository now: git://raeburn.scripts.mit.edu/guile-
emacs.git . (Note: At some point, Emacs will switch from cvs to bzr,
the git mirror will have a new, re-converted history with new commit
ids, and I'll need to rebase everything, which may cause a hiccup if
you're doing anything more than checking out and building.) The
general state of the work is described in README.GUILE at the top
level, also at http://www.mit.edu/~raeburn/README.GUILE.txt .
If you play with it a bit, please let me know -- if it works, if it
breaks, if you have patches...
I think the biggest obstacles on my side to converging with Daniel
Kraft's work are getting the string and symbol representations
converted to native Guile ones from smobs. Both objects currently
carry data in Lisp that they don't in Scheme. Anyone want to help? :-)
Ken
P.S. I was noticing recently that my first commit to the Emacs source
tree relating to this project was on 1999-08-28, just shy of ten years
ago. I'm not sure if I should celebrate or be depressed on Friday. :-)
P.P.S. It looks like work on Emacs 24 may be starting sooner rather
than later. Might be harder to get this wrapped up in time and make a
case for including it in 24...