Re: Success with Mac OS X/CMUCL

Christopher Connolly said:
> Thanks to everyone who worked on getting the Mac OS X port going - I
> downloaded the binaries from Pierre's website yesterday to try out.
> I'm working with Lynn Quam on a system that makes heavy use of external
> C/C++ libraries, OpenGL and Tcl/Tk, all called from Lisp with callbacks
> from C back to Lisp.  Except for some minor trouble with endian-ness
> (tag bits in foreign array allocation), it was very easy to get the
> system running under Mac OS X/CMUCL. We're currently using Lynn's
> callback code as a fallback in lieu of Helmut Eller's system, but so
> far so good.

Good to hear!

> We have the same system running on Mac OS X/Allegro, and I'm noticing
> that while performance favors CMUCL on the x86/linux versions, this
> initial port of CMUCL is slower than allegro on the Mac - not by a
> whole lot, but it's noticeable.  Nonetheless, having a Mac OS X/CMUCL
> is very good news - thanks again!!

While this might be a x86 is special situation, I think the PPC backend is
likely to be a bit less tuned than the x86 or SPARC backends... I'm
looking at going over the backend in the future with an eye towards both
optimizing and fixing it in places that could use attention.

Regs, Pierre.

-- 
Pierre R. Mai <pmai@xxxxxxx>                    http://www.pmsf.de/pmai/
 The most likely way for the world to be destroyed, most experts agree,
 is by accident. That's where we come in; we're computer professionals.
 We cause accidents.                           -- Nathaniel Borenstein