dev meeting 20220217

Present at the meeting:

• Marek Kubica (@Leonidas-from-XIV)
• Nathan Rebours (@NathanReb)

Meeting notes

We discussed the hybrid solver, agreed to go for a solution that will work in most cases and spits out a coherent lockfile, with everything pinned down to a version. Our goal is to make sure that adding opam-provided packages to the equation won't alter the duniverse.

We talked about the --prefer-cross-compile. Marek initially suggested to make it more generic but agreed to stick to this for now and generalize it if needed. It's fairly specific so this scenario is unlikely. Having a specific flag makes it also easier to document its behaviour and to use since one does not have to wonder how to configure it.

We talked about the vendor variable bug. In its current form it cannot be interpreted by opam. We need a short term fix and a long term one. Nathan will reach out to the opam devs to figure out the best solution here.

Nathan worked on tezos, rebasing the opam-monorepo build again. Among the upstream changes to tezos, they upgraded their version of hacl-star so he had to upgrade the hacl-star port. We now have a new lockfile!

Marek worked on improving duniverse-lint further but most of the work he did on opam-monorepo was reviewing PRs and addressing reviews on his.