It looks to me that, as of last night, the bigenough
package no longer builds against the current Coq master branch:
Error: The compilation of coq-mathcomp-bigenough failed at "/usr/bin/make -j 2".
[...]
# File "./bigenough.v", line 87, characters 46-52:
# Error: Syntax error: [ident] expected after 'in' (in [hloc]).
any chance to merge this (seemingly simple) PR soon? Would be nice to have opam archive CI working for a bunch of packages in extra-dev
(including a current PR to the Coq opam archive): https://github.com/math-comp/bigenough/pull/5
cc: @Cyril Cohen @Kazuhiko Sakaguchi
Karl Palmskog said:
any chance to merge this (seemingly simple) PR soon? Would be nice to have opam archive CI working for a bunch of packages in
extra-dev
(including a current PR to the Coq opam archive): https://github.com/math-comp/bigenough/pull/5cc: Cyril Cohen Kazuhiko Sakaguchi
Hi, ... I'm pretty reluctant to merge without CI, I will merge as soon as a CI is added and passes
@Cyril Cohen I just added a Github action file to the PR, but you may have to do something to activate it : https://github.com/math-comp/bigenough/pull/5
I have to admit I don't understand the problem with meta.yml
brought up in https://github.com/coq/opam-coq-archive/pull/1975 -- everything seems to look good to me in the bigenough repo. However, the opam archive PR got stuck in linting for an old opam file that was modified.
I don't think this was the point raised by Cyril, but rather that the 1.0.0 version had:
build: [ make "-j" "%{jobs}%" "COQEXTRAFLAGS+=-native-compiler yes" {coq-native:installed & coq:version < "8.13~" } ]
ah, then I understand. I think this can actually be handled by the current templates, the variable make_target
.
I don't have time to do a PR now though, maybe later tonight.
Last updated: Oct 13 2024 at 01:02 UTC