Closed yutakang closed 11 months ago
An example for this is qrevflat var_0 nil2 = var_1 ==> qrevflat var_0 (x var_2 nil2) = x var_1 var_2
in Prop-28.
Probably these harmful conjectures are produced by proof by induction followed by some explicit conjecturing.
This was done a while ago.
Since we are not using the
cases
tactic, thecase_tac
tactic, or any explicit conjecturing that introduces meta-assumptions, when a conjecture has a meta-assumption that is equivalent toFalse
even though the original goal does not contain a meta-assumption that is equivalent toFalse
, then the conjecture is bad.