Deprecate elimtype and casetype tactics#16904
Deprecate elimtype and casetype tactics#16904coqbot-app[bot] merged 4 commits intorocq-prover:masterfrom
Conversation
|
The job library:ci-fiat_crypto_legacy has failed in allow failure mode |
|
A cursory look at the sources of the broken CI devs seem to indicate that all uses of |
This means turning some elimtype False into exfalso.
|
The use of |
Adapt w.r.t. rocq-prover/rocq#16904 by removing references to "elimtype False".
(cherry picked from commit de18c89)
|
The job library:ci-fiat_crypto_legacy has failed in allow failure mode |
We take the opportunity to use the stdlib-registered empty type instead of the literal Coq.Logic.False type, so as to be more modular.
f6f2638 to
dd25f5b
Compare
Adapt w.r.t. rocq-prover/rocq#16904.
Adapt w.r.t. rocq-prover/rocq#16904. See merge request flocq/flocq!20
For some reason I only commited the elimtype removal in the MR, and didn't push the casetype removal.
Adapt w.r.t. rocq-prover/rocq#16904 (again). See merge request flocq/flocq!21
And stop manually adding it in common.edit_mlg.
|
I have pushed a commit fixing the issue I had reported in the doc, as well as |
|
@coqbot merge now |
There already is such a check, but it is only applied when you add the production to a nonterminal. It's not applied for editing operations such as |
These are basically vestigial tactics that survived the V7-V8 geological crisis. They only survived in some rare instances in the stdlib, and I suspect that they're just not used in the CI. In order to test this the last commit of the PR removes them just to see what happens, as an experiment.
EDIT: all the instances of elimtype removed from the stdlib in this PR were already there in Coq V5, so we're really talking archaeology here.