With all due respect, it appears that I need to reframe my thinking about
these YACC parser builders and go backward in time by about 30-40 years. Now
that I'm getting the hang of what it is doing, it looks strongly reminiscent
of the style of programming we did back on old IBM 360's.

I do notice that YACC becomes very sensitive to slight perturbations in the
grammar. This is not meant as a criticism of YACC, but rather a plea for
some more modern thinking in compiler generators. Perhaps the GLC tool will
do better here.

The problem seems to be the impedance mismatch between the way humans are
led to think when viewing the "structured" YACC input clauses, and the more
or less blind way the computer is actually using that information. I see
that being a master of removing reduce/reduce conflicts seems to be a rite
of passage. But surely there has to be a better way to approach these kinds
of problems. Just because my ancestors used to hunt with spears doesn't mean
I still ought to do so.

David McClain
Senior Corporate Scientist
Avisere, Inc.

+1.520.390.7738 (USA)
david.mcclain@avisere.com



-------------------
To unsubscribe, mail caml-list-request@inria.fr Archives: http://caml.inria.fr
Bug reports: http://caml.inria.fr/bin/caml-bugs FAQ: http://caml.inria.fr/FAQ/
Beginner's list: http://groups.yahoo.com/group/ocaml_beginners