-
-
Notifications
You must be signed in to change notification settings - Fork 3.6k
Build: incompatible Mercurial versions from VCS setup than build commands #8995
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
Comments
I only skimmed through the logs, but this is weird. It looks like the So, might this is no longer the same hg and some of the virtualenv play in between downgraded it somehow ? |
I don't have any insight into reathedocs infrastructure, though a co-worker pointed out a somewhat suspicious commit to Mercurial which would have been released a couple weeks before this started failing. But yeah, it's weird that the earlier Edit: Apparently marmoute ruled out this commit in conversation with said co-worker. I'll leave this comment up just in case. |
The "suspicious commit" (and the related work in this area) only affects the "streaming clone" feature. Streaming clone does not seems to be involved in this build at all. So I would be surprised if these changes play any part in the current failures. |
We are seeing this on the PyPy doc builds as well. Is there a way to find all the projects that are using mercurial to find one that is succeeding? |
According to #8840 there are ~46 repos using mercurial. Are all their builds failing? |
@ahal can you try using a Read the Docs config file with build.os (https://docs.readthedocs.io/en/stable/config-file/v2.html#build-os) declared as Ubuntu 20.04? |
I guess the problem here is that we are using a newer version in the I think you will be able to fix this problem by using Ubuntu 20.04 in your config file as I suggested in my previous comment. It would be good to find a way to tell Mercurial to clone the repository to make it compatible with older versions. |
You can use |
Thanks, this works around the issue for me. |
We are using Ubuntu 20.04 LTS to perform the clone step (`readthedocs/build:ubuntu-20.04`) which comes with Mercurial v5.3.1 and install by default the `sparse-revlog` requirement. However, if the build is using `readthedocs/build:latest`, which comes with Mercurial v4.5.3 and does not contain that requirement it fails: ``` hg identify --id abort: repository requires features unknown to this Mercurial: sparserevlog! (see https://mercurial-scm.org/wiki/MissingRequirement for more information) ``` This commit disables that requirements to make the clone compatible with the older version of Mercurial. Closes #8995
We are using Ubuntu 20.04 LTS to perform the clone step (`readthedocs/build:ubuntu-20.04`) which comes with Mercurial v5.3.1 and install by default the `sparse-revlog` requirement. However, if the build is using `readthedocs/build:latest`, which comes with Mercurial v4.5.3 and does not contain that requirement it fails: ``` hg identify --id abort: repository requires features unknown to this Mercurial: sparserevlog! (see https://mercurial-scm.org/wiki/MissingRequirement for more information) ``` This commit disables that requirements to make the clone compatible with the older version of Mercurial. Closes #8995
Something is wrong. I cannot build the PyPy docs https://readthedocs.org/projects/pypy/builds/ for the last day: the build fails to get |
@mattip for some reason |
Thanks for looking. |
Do you have any error or traceback? We could look at ? |
@marmoute if you mean me, I only have the information in the build log https://readthedocs.org/projects/pypy/builds/16514248/ |
I've already found the issue and opened a PR to fix it at #9055 --It's a Read the Docs' bug, not related with Mercurial itself. |
This is now deployed. Please let me know if this solves the issue. |
Looks like pypy is working properly: https://readthedocs.org/projects/pypy/builds/16518308/ |
…bug) This is an attempt to work around: readthedocs/readthedocs.org#8995
Details
Expected Result
The builds should succeed :)
Actual Result
Builds fail as seen here. I don't believe anything changed on the repo side, this was working a week or two ago.
The text was updated successfully, but these errors were encountered: