Open andrew-appel opened 1 month ago
Duplicate of #18
I'm not going to fix it very soon. I suggest using the set
tactic as a workaround.
This bug is also related to a (Coq) feature we recently discussed with @CohenCyril and @Tragicus, that would eagerly simplify (GRing.one Datatypes_nat__canonical__GRing_SemiRing)
to (S O)
at goal "creation time". With that in place, ring would work as expected.
@gares I think GRing.one Datatypes_nat__canonical__GRing_SemiRing
should not be simplified to S O
automatically. This issue has to be eventually fixed in Algebra Tactics.
In this example, we see that the
ring
tactic fails or succeeds on terms that are equivalence modulochange
, and in fact the original term comes up naturally in one of my proofs. Is this a bug or a feature?And here is the detailed version information: