`Equivalence` for `≡` on gmultisets.

theories/gmultiset.v

@@ -45,7 +45,7 @@ Section definitions.
 
   Global Instance gmultiset_dom : Dom (gmultiset A) (gset A) := λ X,
     let (X) := X in dom _ X.
-End definitions.
+End definitions. 
 
 Typeclasses Opaque gmultiset_elem_of gmultiset_subseteq.
 Typeclasses Opaque gmultiset_elements gmultiset_size gmultiset_empty.
@@ -66,6 +66,8 @@ Proof.
 Qed.
 Global Instance gmultiset_leibniz : LeibnizEquiv (gmultiset A).
 Proof. intros X Y. by rewrite gmultiset_eq. Qed.
+Global Instance gmultiset_equivalence : Equivalence (≡@{gmultiset A}).
+Proof. constructor; repeat intro; naive_solver. Qed.
 
 (* Multiplicity *)
 Lemma multiplicity_empty x : multiplicity x ∅ = 0.

[Ralf Jung]

Do we need a CI linter rejecting trailing spaces? ;)

[Robbert Krebbers]

Or a git pre-commit hook?

[Ralf Jung]

That's something you'd have to do locally. I could probably to a push hook, but I am not sure if GitLab exposes a nice interface for that.

[Robbert Krebbers]

By the way, in case we do anything like this, it should just check that no trailing spaces have been added, not that there are no changes to spaces at all.

Sometimes we want to fix badly spaced stuff, ofc, and that should be accepted.

[Ralf Jung]

I thought it'd check that there are no trailing spaces anywhere in the repo after the commit.

I also was slightly joking^^ but we could have a linter CI pass, probably before the main build jobs that take a while. Linting should take essentially zero time.