NeuroTechX / moabb

Mother of All BCI Benchmarks
https://neurotechx.github.io/moabb/
BSD 3-Clause "New" or "Revised" License
678 stars 176 forks source link

MAINT: Try not using poetry env #578

Closed larsoner closed 5 months ago

larsoner commented 5 months ago

From MNE-Python office hours, just a quick one-off idea on how to solve an env problem. Not sure if it'll work but let's try!

Also adds a concurrency to cancel in case multiple pushes are done. Maybe you don't want this but I thought I'd put it in in case you do.

larsoner commented 5 months ago

Needs a dev approval to run

bruAristimunha commented 5 months ago

Hi @larsoner,

Sorry for the not related issue. It is fixed now.

larsoner commented 5 months ago

Can you point me to the failing step in your CIs? I assumed it was with the "Building Docs" step but it looks like that was passing on the latest commit(s) to develop

bruAristimunha commented 5 months ago

Yes @larsoner! During the deployment step.

We have two deployments on the page, one of which is done manually by moving the entire building to the github repository (https://github.com/NeuroTechX/moabb.github.io). Another one is done directly with GitHub action. Both are not working.

Here:

https://github.com/NeuroTechX/moabb/blob/72a530b2f614999e4ad79ea2d26b841497969b0e/.github/workflows/docs.yml#L135-L144

bruAristimunha commented 5 months ago

Here, @larsoner https://github.com/NeuroTechX/moabb/actions/runs/8853163014/job/24314254029

larsoner commented 5 months ago

A bit tough to work on without the actions running etc. I could open a tiny PR or something but not sure how much time I could spend looking into it at the moment anyway. So in general I'd suggest adding stuff like ls -al etc. in the run and test ... type statements to make sure things are where you think they are. That plus commenting out some code (like I did in the latest commit commenting out an if:) to make sure the steps actually run in a PR can be helpful. In theory as long as you open a PR from a fork (rather than a branch) it shouldn't actually push, but in practice I usually comment out whatever could do damage (like git push origin gh-pages) to be safe. Good luck!