Skip to content

Conversation

@AayushSabharwal
Copy link
Member

@AayushSabharwal AayushSabharwal commented Jan 7, 2026

Close #4211

@AayushSabharwal
Copy link
Member Author

I'm consistently stumped by how this manages to have different behavior on LTS, and only in CI.

@ChrisRackauckas
Copy link
Member

LTS doesn't obey sources

@AayushSabharwal
Copy link
Member Author

Ah, right thanks. I thought I made the CI script dev packages appropriately, but apparently not.

@AayushSabharwal
Copy link
Member Author

Hmm, but that doesn't explain the ModelingToolkitBase/InterfaceI failure. There's no sources entry in that package.

@AayushSabharwal
Copy link
Member Author

Okay yeah, that isn't it. I made the runtests.jl file dev the appropriate subpackages. Still, I think it's prudent to do it in the CI script instead. The Pkg.status printout is then always correct, and we avoid stacked environments.

@AayushSabharwal AayushSabharwal force-pushed the as/fix-more-tests branch 4 times, most recently from 1e367f3 to bdc1eba Compare January 8, 2026 15:40
@AayushSabharwal
Copy link
Member Author

The print statements seem to affect tests passing. In fact, they affect completely unrelated tests.

@AayushSabharwal AayushSabharwal force-pushed the as/fix-more-tests branch 3 times, most recently from cdc0da3 to f5dffa9 Compare January 13, 2026 08:02
@AayushSabharwal
Copy link
Member Author

Hopefully the suggestion in #4211 fixes the stochasticity here

@AayushSabharwal
Copy link
Member Author

It doesn't look like the fix worked

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

collect_vars! bug on 1.10 when loading DynamicQuantities

3 participants