Closed pi8027 closed 3 years ago
Looks sensible, any reason this is marked as a draft?
Also, do you have an estimate of which version of mathcomp the deprecation will take place in, will that be 1.12 already?
Looks sensible, any reason this is marked as a draft?
Since I finished documentation work for math-comp/math-comp#601 (but I'm still waiting for review there), I marked this PR as ready for review.
Also, do you have an estimate of which version of mathcomp the deprecation will take place in, will that be 1.12 already?
Since now it targets 1.12, I hope it will be included in 1.12.
Since now it targets 1.12, I hope it will be included in 1.12.
Hi @pi8027. Let me clarify the meaning of deprecation in your answer: do you mean that you just intend to merge #601 for the 1.12 milestone, or that, once it is merged, you also expect to remove the corresponding deprecate
notations, e.g.:
Notation "@ 'eq_sorted_lt'" := (deprecate eq_sorted_lt sorted_eq_lt)
(at level 10, only parsing) : fun_scope.
for the same release milestone?
LGTM!
@germanD No. These deprecation notations such as "@ 'eq_sorted_lt'"
will be included in the 1.12 release so that libraries depending on MathComp can support both ~1.11 and 1.12. Then it will be removed at some point between 1.12 and 1.13.
@pi8027 thank you!
Indeed, thanks @pi8027!
We are about to rename
eq_sorted
lemmas tosorted_eq
to address a naming inconsistency in MathComp (see https://github.com/math-comp/math-comp/pull/601#issuecomment-705046342). Unfortunately, this deprecation breaks fcsl-pcm and lemma-overloading because the deprecation facility of MathComp does not support the(ident := term)
syntax for explicit applications. As a workaround, I propose to use@
syntax instead.