Open epatters opened 1 year ago
Yeah it looks like there was a breaking change in a dependency. I was looking into this earlier today actually. Looks like the error is coming from Symbolics.jl which is an indirect dependency
Thank god for "semantic versioning" allowing us to update freely without hitting breaking changes...
Also for reference, here is a force re-run of the tests a few hours ago if anyone wants to look at the logs:
https://github.com/AlgebraicJulia/AlgebraicPetri.jl/actions/runs/6656621641
julia> import OrdinaryDiffEq, ModelingToolkit
julia> OrdinaryDiffEq.ODEProblem === ModelingToolkit.ODEProblem
true
Is this equality a recent change? It would explain why the extensions are stomping on each other.
I actually pulled out an old laptop and can run the current main
branch successfully :joy: thank god for not updating old laptops LOL
I am going to investigate the current versions and see if I can identify which version of something caused the break
Also just tested in this setup and that equality is holding true even in the environment I have where it works. I will try to identify what version of Symbolics/Catalyst/Julia breaks this
Looks like updating all dependencies except pinning Symbolics to 5.5.0 it works. Updating just Symbolics to 5.10.0 introduces the issue. Currently bisecting the problematic version, but it's taking some time... dang precompilation
Oh man, I just realized I misread the original issue here and am talking about a different issue...
Sorry for the noise here! I marked my previous comments all off topic
It looks like this severely stalls the precompilation of SimulationService.jl. I get a fun mix of warnings and errors. Happy to share the terminal output if it will help.
Closed by #167 as soon as tests pass!
When precompiling the OrdinaryDiffEq and ModelingToolkit extensions, I get the following warnings:
This is on Catlab v0.15.5.