On Nov 15, 2004, at 4:49 PM, Raymond Toy wrote:
> mikel evins wrote: >> That would certainly be very useful to us. Carl said something about >> doing a Unicode implementation based on some Lucid code; are those >> separate codebases? > > The old implementation was donated by Brian Spilsbury (sp?), so this > is probably separate. It's certainly news to me since the only > implementation I've heard about was by Brian. > > Bringing it up-to-date very likely won't be easy. It might be easier > to start over from the current sources. I suspect Brian had a hard > time doing it. Pierre Mai did some nice work, I think, to get Brian's > version ported to the current version at that time. Another issue that is bothering us some is multiprocessing. We maybe able to use the CMUCL implementation on x86, but if I understand correctly, it's not available on PPC. That might cause us to use OpenMCL on PPC. (Maybe it would be simpler to use a single Lisp for all targets, but, on the other hadn, one could argue that using two of them will aid portability of the sources.) --me
