]> matita.cs.unibo.it Git - helm.git/commit
- GREAT: when unifying ?1 : Type[i] with ?2: Type[j] the unifier
authorAndrea Asperti <andrea.asperti@unibo.it>
Fri, 26 Nov 2010 13:18:50 +0000 (13:18 +0000)
committerAndrea Asperti <andrea.asperti@unibo.it>
Fri, 26 Nov 2010 13:18:50 +0000 (13:18 +0000)
commit73b49101cb8f83bc54fd2dc3c862b42c00ad13b5
tree0861406e9d0cfb805fa55a51d89c90f81a80b84b
parent48a389fa5b4c7d80f97c8c6c7a1f47e840977f39
- GREAT: when unifying ?1 : Type[i]  with ?2: Type[j] the unifier
  randomly tried to instantiate ?1 with ?2 even when j > i, yielding to
  an unification error later on. In turn, this created that horrible
  behaviour of rewriting that failed if you did not pass enough types
  to the lemma.
- in rewriting the argument is now automatically saturated

These two fixes together allow most of the time to write simply >f as we
did in the old system.
matita/components/ng_refiner/nCicUnification.ml
matita/components/ng_tactics/nTactics.ml