On 6/9/07, Richard Fateman <fateman at eecs.berkeley.edu> wrote:
> argument against:
> slows down simplifier.
If the concept is useful, we can code in Lisp and it would be faster.
Maybe all we need to do is adjust the existing "threadable" stuff.
But I'm not in any hurry to do that; since the simplification rule is
attached to new operators (built-in operators already have suitable
distributes-over stuff hard-wired into their simplifiers or elsewhere),
this distributes-over scheme won't slow down simplification of
built-in operators. So I'd like to get the user-visible behavior
straight first, then we can quibble over the implementation.
(Ideally all operators would have their properties expressed by
the same mechanism, but we're a long way from that and I'm
not ready to tackle it.)
FWIW
Robert