Closed Dr-Irv closed 2 years ago
Hi, this needs a minimal reproduction to be addressable -- it is highly unlikely that anyone will find the time or interest to debug the CI pipeline of a different project. Could you please create a minimal reproduction (preferably, but not required to be in a Windows container)?
besides which, the failures that you are attributing to poetry 1.2 are happening with your CI using poetry 1.1.15 eg https://github.com/pandas-dev/pandas-stubs/runs/8105323908?check_suite_focus=true
So you'll want to dig a bit deeper yourself I reckon
Hi, this needs a minimal reproduction to be addressable -- it is highly unlikely that anyone will find the time or interest to debug the CI pipeline of a different project. Could you please create a minimal reproduction (preferably, but not required to be in a Windows container)?
That will take a bit of work on our part to isolate. It could be only in CI, or maybe regular Windows, or due to combinations of versions of software. I'll see what we can do. What I do know is that pinning poetry back to 1.1.5 makes our CI work.
besides which, the failures that you are attributing to poetry 1.2 are happening with your CI using poetry 1.1.15 eg https://github.com/pandas-dev/pandas-stubs/runs/8105323908?check_suite_focus=true
So you'll want to dig a bit deeper yourself I reckon
No, that run was failing for other reasons.
If you look at the successful run here: https://github.com/pandas-dev/pandas-stubs/actions/runs/2958394397 versus the unsuccessful run here: https://github.com/pandas-dev/pandas-stubs/actions/runs/2962734028
on the same PR, the difference in the logs revealed poetry 1.1.15 worked and poetry 1.2.0 did not.
well one of us is having a reading comprehension problem, maybe it's me.
How is the run that I pointed at different? Looks exactly like the windows-only loguru stuff in the other one to me.
(You keep typing 1.1.5 - can you really mean that or is 1.1.15 intended?)
well one of us is having a reading comprehension problem, maybe it's me.
How is the run that I pointed at different? Looks exactly like the windows-only loguru stuff in the other one to me.
Fair enough. I was concentrating on the PRs that were working and started failing after the upgrade to poetry. Can't say what happened on that particular one that you found.
(You keep typing 1.1.5 - can you really mean that or is 1.1.15 intended?)
Ugh, I keep making the same typo!!! I meant 1.1.15.
So I also don't know what happened here but if we agree that the attribution to poetry 1.2.0 is faulty then probably this issue should be closed.
(I guess you hit something transient or cache-y or something on github, and coincidentally it fixed itself when you changed poetry version. But as neersighted suggests, I don't really care to investigate...)
So I also don't know what happened here but if we agree that the attribution to poetry 1.2.0 is faulty then probably this issue should be closed.
I'm not sure that I agree that the attribution to poetry is faulty.....
well you agree that the same thing happened at poetry 1.1.15, which makes it pretty darned suspect
And I just replicated it with poetry 1.1.15, so I guess that wasn't it. I think it has to do with how we are caching poetry.lock
Thanks for finding that it wasn't working with 1.1.15
This issue has been automatically locked since there has not been any recent activity after it was closed. Please open a new issue for related bugs.
[x] I am on the latest Poetry version. (version 1.2.0)
[x] I have searched the issues of this repo and believe that this is not a duplicate.
[N/A] If an exception occurs when executing a command, I executed it again in debug mode (
-vvv
option).OS version and name: Windows Server 2002
Poetry version: 1.2.0
Link of a Gist with the contents of your pyproject.toml file: https://github.com/pandas-dev/pandas-stubs/pull/241/files
Issue
We had a pretty simple CI running on Github for the pandas-stubs project. It started to fail on Windows, because it started pulling in version 1.2.0 of poetry.
When I pinned the poetry version to 1.1.5, our CI worked fine. The linked pull request shows that change, and you can visit https://github.com/pandas-dev/pandas-stubs/actions to see what we are seeing in the logs of failed runs, which were failing because a package couldn't be found that poetry said it installed.