-
Notifications
You must be signed in to change notification settings - Fork 19
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
Make 0install-solver available in the opam 2.1 installation #102
Conversation
Looks like this depends on a commit that isn't merged yet. |
yes, it's on purpose, I remember @avsm saying starting with docker-base-images is better to avoid issues, at the last meeting |
Yep, that's right. It's best to ensure things build here and then merge them (immediately) since we dont have a CI-for-the-CI. I'm just doing the multicore changes and then will look at this... |
@kit-ty-kate wouldn't this be better as a patch to opam itself? I'm a little unclear on what's going on here -- we install the external 0install solver, and then it becomes available to opam as builtin-0install? |
38a5892
to
cbcb5bf
Compare
The required PR in opam has now been merged so this PR is now ready to be reviewed. |
The ocaml-dockerfile PR needs rebasing on master before we can push this to |
6494260
to
1da0e66
Compare
Done but I'm not sure why Github is still saying that there is a conflict :/ the |
1da0e66
to
71ac3b2
Compare
Oh I see, I was only looking at Hopefully it should be ready now, sorry for the setbacks. |
Deployed to |
mmh, I think this just hit something akin to #104 :/ |
Hmm, yes, this line needs to include the hash of the commit it wants. |
c02ac0d
to
6dfaf44
Compare
This PR is ready for another round of review. Incidently it should also partly fix #104 |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
This is a good change, but I think there is a small problem: it will rebuild all the images every time one of the opam branches or opam-repository updates.
Having everything on the weekly schedule could help if they all line up, but they won't in general (and if someone manually rebuilds one of them they'll get out-of-sync again).
We should probably have the three commit inputs stay up-to-date, but have them all go into a snapshot box that takes a snapshot of them once a week.
I spent the past week and a half trying various patches to ocurrent to make the So in the meantime this gets resolved, I used a single cache entry to assemble the three of them in this particular context. |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
This seems reasonable to get things unblocked. I've pushed it to live
.
Doesn't work, though:
https://base-images.ocamllabs.io/job/2021-05-10/144920-git-repositories-01ff68 |
oops, sorry for that. It's fixed. |
Still doesn't work:
Does it work for you locally when you test it with |
…citly pass the git hash for the opam source to keep all the images consistant
62d5e61
to
c7d9ded
Compare
Sorry for that, I really have no disk space to give to docker on my local machine and until now I didn't know my dev work server had its ports open. I had never had the need to test docker-base-images locally in the past so i didn't know how to test it either. Now i know, sorry for that. I've now tested that at least the default alpine image is built all the way and contains opam 2.1 with the 0install solver enabled. |
OK, seems to be working now! Just needs ocurrent/ocaml-dockerfile#43 merging first. It seems to take a long time to get to 0install though! $ opam-2.1 install --dry-run --debug --solver=builtin-0install utop
...
00:07.351 0install Solution found. Solve took 0.31 s |
Merged the dockerfile pr. Isn't a chunk of that time opam-2.1 doing a format upgrade on ~/.opam from 2.0, @talex5? |
I thought I ran that command multiple times and it was always slow. |
yeah it's not the upgrade to 2.1 (with the upgrade it take 17s to get to it), most of the bulk of the time is spent in the transformation to CUDF (4s), untaring the repository (2s) and detecting the depexts (1s):
|
No description provided.