1/ Imagine we work to solve this programming problem: < f z : T > where z :
Sigma X Y.
We would like to have "eta-rules" while pattern-matching, that is the ability
to write:
f (x , y) instead of invoking the split eliminator.
2/ Further, when working on <flip g : B -> A -> C> with g : A -> B -> C, it
would be exciting to be able to write:
flip (\ x y -> f y x) = f
(1) seems a reasonable target, (2) is more extravagant.
Original issue reported on code.google.com by pedag...@gmail.com on 23 Aug 2010 at 1:44
Original issue reported on code.google.com by
pedag...@gmail.com
on 23 Aug 2010 at 1:44