Eli Barzilay  <eli@barzilay.org> wrote:
+---------------
| ...many concepts that are deep in the core of Lisp (e.g, packages
| instead of modules) turned out to have better solutions than these
| in Lisp, which makes Scheme implementation diverge on them instead
| of converging back to Lisp.
+---------------

In particular (since MzScheme *was* mentioned), the PLT groups's[*] work
on first-class compilation "units" (modules) seems quite significant
[see <URL:http://www.cs.rice.edu/CS/PLT/packages/doc/mzscheme/node54.htm>
for technical details], and is backed up with both theoretical research
<URL:http://www.cs.rice.edu/CS/PLT/Publications/index.shtml#papers> and
several years of practical experience in implementing large systems with it
[the whole DrScheme/MrEd/MzScheme layering is based on units-with-signatures].

It would be interesting to see if there would be any benefit to adding
a similar scheme (no pun intended) to Common Lisp. [Or conversely, whether
there is any point in trying to extend MzScheme's single-inheritance object
system to multiple-inheritance.]

Or perhaps these two worlds are inherently incompatible with nothing
to learn from each other (which, if true, would be sad, IMHO).


-Rob

[*] I use the plural possessive here, since with the graduation
    of several of their PhD's, the original Rice University PLT
    (Programming Languages Team) has spawned additional "colonies"
    at Brown & Utah (see <URL:http://www.cs.rice.edu/CS/PLT/>).

-----
Rob Warnock, 31-2-510		rpw3@sgi.com
SGI Network Engineering		http://reality.sgi.com/rpw3/
1600 Amphitheatre Pkwy.		Phone: 650-933-1673
Mountain View, CA  94043	PP-ASEL-IA