-
Notifications
You must be signed in to change notification settings - Fork 142
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Prepare for Agda 2.6.3 #948
Conversation
The unresolved metas may have to do with the fix for agda/agda#6219 ("may" because I haven't looked: I promised myself I'd take today off... sigh). Sadly we don't yet understand the interaction between injectivity analysis and HITs, so disallowing it is the best I can do for now. |
68a3538
to
c00c032
Compare
@aljungstrom can you fill in the implicit arguments in CupProductTensor that agda 2.6.3 cannot figure out? |
@felixwellen lol did I do it right? I suck at github you know.. made a PR: #953 |
Thanks for looking into it. Seems good to me gitwise. If you have the rights (you do if you are a member of the agda github space), you can just directly push commits to |
Not sure if I really see all the members of the agda space, but if I do, you are not a member... So I guess making a PR on my PR was the best you could do. |
e79f3ac
to
76e3be2
Compare
Now we are prepared for the new agda release - once it is out we can switch to the tag 'v2.6.3' (now it is 'release-2.6.3'). |
Questions from the outside:
|
I'm not entirely sure what to do, but I guess the following makes sense:
@mortberg : Does that sound reasonable to you? |
Yes, this sounds good to me. Thanks for taking care of making things working with the new release! I'm not sure, but maybe your checklist should be added to https://github.com/agda/cubical/blob/master/RELEASE.md ? Does one have to do anything for the NIX stuff? |
I'll merge the nix stuff now, so it will be in cubical-0.4. |
releasing is now #956 |
76e3be2
to
8574a52
Compare
@guilhermehas I'm a bit confused about the behaviour of ci-nix. Why doesn't it check this branch (which I just rebased)? If I guess correctly, it will check once I merge this PR. And then it will fail as long as nixpkgs is still on agda 2.6.2.2... |
You are right. It just goes to CI if it merged into master or if you change any nix or lock file in some pull request. In addition, I have seen that Agda already has |
You can use the It kind of makes sense that @felixwellen 's force push didn't trigger the ci-nix, since none of the new commits touch |
Is this to avoid the ci job failing until agda 2.6.3 is in nixpkgs, or do you think it is generally better to use the agda flake in the cubical flake instead of using the agda package from nixpkgs? |
My idea was to avoid ci job failing, but I don't know which of these two options @felixwellen prefers. |
I don't have an opinion on that. If agda 2.6.3 is out and it takes to long for that to get into nixpkgs, I might just deactivate the ci job. Once agda 2.6.3 is in nixpkgs, we can switch it on again. |
Umm... the flake.lock in this branch already contains the new "rev" for nixpkgs, since we did that in 95499d9. I tried |
But I can push a commit with the |
Yes, please! |
Done, but it does not work in this PR, because "To trigger the workflow_dispatch event, your workflow must be in the default branch." as I now finally found in the github docs. |
But the nix-ci workflow does indeed fail, I got it to run here: MatthiasHu#1 |
I made a pull request compile using Nix and Agda 2.6.3. |
So, to summarize: Once agda 2.6.3 is out, we have to
|
This reverts commit c00c032.
maybe just to see what happens...
Added agda-2.6.3 in flakes
Nixpkgs of Agda using following another nixpkgs
4911e5c
to
ef9d222
Compare
The PR was already discussed -> merging. |
Awesome, thanks to the merge the Nix package for cubical does not need to manually pull this PR in any longer. :-) |
So far, this is only an update of jesper's #921, which just renames --experimental-lossy-unification to --lossy-unification. I updated to see, if there are problems with the upcoming new agda release - turns out there are unresolved metas with the current master of agda (e.g. in Cubical/Homotopy/EilenbergMacLane/CupProductTensor.agda).