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
Re-run CI #203
Re-run CI #203
Conversation
Oh, wow, thanks for catching this. It looks like 4 errors have started to happen:
So I'll keep looking into this last one, and I may work on that flaky convergence one as well. |
Fastapi is a possible culprit here, though they have a really fast release pace so maybe it's nothing. https://anaconda.org/conda-forge/fastapi/files |
Here's the most recent nightly CI run that passed: https://github.com/openforcefield/openff-bespokefit/actions/runs/3231103232/jobs/5290258911 |
Looks like that was it. Possible even rootier root issue is the fastapi conda package changing from requiring starlette 0.20 to 0.21. Rabbit hole begins here: tiangolo/fastapi#5646 |
That's exactly it; the currently-failing tests ( FastAPI is known for being relatively unstable as a framework, which may or may not be fair since Django and Flask are a decade or so older and more hardened. Let's see a different pin helps ... |
^ above failures vs. passing tests is because I forgot to update both environments, not necessarily because of some mystical relationship between OpenEye and Starlette |
If this pin works, my current plan is:
I can also push the pin into the conda builds if you think that's worthwhile |
That plan sounds good to me, and I agree we should pin the conda builds as well! I'd like to gain a better understanding of where the rift in our fitting stack is going to open up - There's some range of packages that work with OFFTK 0.10, and another that works with OFFTK 0.11. Also I'm a little bothered by the sporadic GeomeTRIC convergence errors, but I'll open a separate issue for that. |
Merging these fixes as part of #202 since these environments are pretty slow to solve (one took 50 minutes) |
CI is failing on feature branches for reasons that are not clear to me. I'm re-running CI here to see if the failures are genuine and on the
main
branch.