-
Notifications
You must be signed in to change notification settings - Fork 27
dev meeting 20220609
Thibaut Mattio edited this page Jun 9, 2022
·
1 revision
Present at the meeting:
- Marek Kubica (@Leonidas-from-XIV)
- Thibaut Mattio (@tmattio)
- Nathan Rebours (@NathanReb)
- Marek contacted maintainers and followed up on the open PRs
- Released dum
- Thierry merged PRs. It's unclear if stdcompt and pyml will use dune as the default workflow, need more discussion.
- The MR has been opened for a few weeks on Tezos repository. It’s working and updating the lock file is going smoothly. Although there’s issues to fix everytime.
- Main next step is updating the CI to use opam-monorepo instead of vanilla opam. Nathan to have a talk with Raphael to get a tour of the CI. Will give a shot at building a base docker image for the deps and tag it with a hash of the lockfile.
- Release opam-monorepo that removes the hack we had to do for the initial MR.
- Also will contain the minimal update feature
- Finish moving to 4.14.
- Worked around the issue and released on opam.
- Need to talk with the CI team about the work they are doing for Tezos.
- They need to use opam-monorepo also
- Need to check if it’s worth it for us to work on the Gitlab CI if they are going to move to the ocurrent CI soon
- Thibaut to schedule a chat with Tim
- Hacl-star
- Status of upstreaming: we need to check with hacl-star maintainers if they’d be open to upstream the dune port in its current state (not the final solution)
- Marek to discuss with the hacl-star team
- Solver error messages are still a painpoint, not clear how to solve this as 0install might be unable to give better errors, but we should look into it.
- Thibaut to setup a meeting with CI team at Tarides to discuss intesections of opam-monorepo and the Tezos CI
- Nathan to update the lockfile to 4.14
- Nathan to work with CI team at Nomadic Labs to port the Tezos Gitlab to use opam-monorepo
- Marek to continue to discuss with upstream package maintainers to port the projects to dune
- Marek to discuss with hacl-star maintainers on how they feel about usptreaming our existing hacl-star port
- (if time allows) Marek to look into new opportunities to improve error messages related to solver failures.