On Tue, 06 Nov 2007 21:29:56 +0900
Kenichi Handa <handa@ni.aist.go.jp> wrote:

> In article <m21wb3vayb.fsf@news.cyberhut.org>, CHENG Gao
> <chenggao@gmail.com> writes:
> 
> > If Carbon port is treated as dead, and emacs-app (Cocoa port,
> > http://emacs-app.sf.net) works well, is it possible that emacs-app
> > be merged (as a branch in cvs or git repo) thus we MacOSX users can
> > have an evolving Emacs?
> 
> To my understanding, Carbon port should be treated as dead,
> and the merging of Cocoa port into the trunk (or to
> emacs-unicode-2 branch) isn't that far.

I surely hope that the Carbon port isn't dead. 
I use 22.1.1 2007-06-16 on powerpc-apple-darwin8.9.0
as the only emacs that works on my laptop.

It is not a subjective sort of "doesn't work", I tried
emacs.app (can't remember version will add more) and it blew up on my .emacs
with these lines:

(cond
  ((string-equal "gnu/linux" system-type) (load-file "/usr/share/emacs/site-lisp/site-gentoo.el"))
  ((string-equal "darwin" system-type) (load-file (concat (getenv "HOME") "/system/emacs/darwin.el")))
  )

For a test case you can drop the (load-file "file") part. I found that a single string-equal call
on the system-type symbol worked. However two subsequent uses of the system-type symbol as above
caused the elisp debugger to be entered.

I removed the line and it (Emacs.app) continued to exhibit terrible behavior. Crashed during syntax highlighting,
received numerous "max binding depth reached" like errors doing innocuous things like C-hf. I have also noticed
that problem on Linux-x86 with emacs 23.0.0.3, but it's alot harder to trigger there.

I was waiting for the time to construct a more useful problem report, but this has at least a test-case.