Re: yet another XEmacs fork;-)
Julian Bradfield writes:
As I wrote in my previous reply to Stephen, but then deleted before
sending, "it's not as if Mule does any of the hard stuff";-) [so why
is it hard to get rid of?]
Compatibility. Also, there *is* a lot of hard stuff in Mule.
Regexps, for one. Font-handling, for another; OOo and Firefox do a
crummy job with mixed Han texts. I don't know how we're going to
emulate Mule in that respect with Unicode as the primary internal
character set. Seriously f*cked-up file systems are another (try
reading a file whose name contains UTF-8, KOI8-R, and Shift JIS in
different segments in them other apps ... yes, I've seen such!) In
fact, Python just went through a big debate on coding systems for the
fiile system, which ended when Guido declared that designing a system
to do it right was too hard, so they went with a 99% proposal. And
there's a huge amount of tweaking in the detection code and other
places.
_______________________________________________
XEmacs-Beta mailing list
XEmacs-Beta(a)xemacs.org
http://calypso.tux.org/cgi-bin/mailman/listinfo/xemacs-beta