That range includes tabs and new lines.

Thanks Morten for spotting this problem.

This way we can use set_solver to solve goals involving ∈.

This has bothered me repeatedly in proofs, now I finally got around to fix it at the source

This reverts commit 2a7755fe because
it is no longer needed after Matthieu Sozeau reverted this change in
Coq 8.6. See also the discussion:

  [Coq-Club] Coq 8.6 typeclasses behavior change

at 11/16/2016 02:14 PM.

We need instances like EqDecision and Countable for it. We could
redeclare those instead, though.

In Coq 8.6 type class search is not called recursively on premises
that are not type classes. To that end, we use a hint extern to invoke
an ordinary auto.

Many useful properties are probably still missing.

(These instances are not defined for any FinMap to avoid overlapping
instances for EqDecision, which may have awkward consequences for
type class search).

There is no way to infer the cmra A, so we make it explicit.

This way we avoid the env_cbv tactic unfolding string related stuff
that appears in the goal and hypotheses of the proof mode.

Having Is_true as a type class caused problems with rewrite: when the
rewrited lemma has a premise of the shape Is_true, the rewrite tactic
will complain that it cannot find a type class instance, instead
of generating a goal for that premise.