Cédric St-Jean

@AayushSabharwal I get the software engineering angle, but from my chemical engineer colleagues' perspective: > MTK has the ability to substitute components. See https://docs.sciml.ai/ModelingToolkit/stable/API/model_building/#ModelingToolkit.substitute_component. The idea is that you have...

Thank you for the answer. > Because changing a component changes the structure of the system, and it needs to be resimplified What is the use-case for post-simplification substitution? I...

Awesome, exactly what we need! If it's in the documentation then we should close this issue, but I couldn't find it here or in Symbolics.jl.

I may have some understanding of the current test failure. It seems that, with ```julia @mtkmodel SimpleLag begin @structural_parameters begin K # Gain end end @mtkmodel DoubleLag begin @parameters begin...

I figured it out (`@named` uses `default_to_parentscope`) and fixed it in https://github.com/SciML/ModelingToolkit.jl/pull/4029/commits/4086714d2cdae37f2f693d2af46d2592dfe30c5a

> This is... interesting. Given that @mtkmodel is being deprecated in v11, I don't mind this as long as it passes tests. It might be worth holding off a bit,...

> @mtkmodel won't be part of MTK in v11, but will be split out into a separate package along with tests. It will still live in this same repo, and...

> It will still live in this same repo Somehow I missed that part. This suggests that it'll still be part of the MTK test suite throughout v11? If so,...

At the very least, `julia` or `sudo julia` should always start julia, regardless of whether the `juliaup` check failed or not...

I am not well-versed in Unix permissions, but on an OSX machine that does not use juliaup (the `julia` executable is in the PATH), `sudo julia` works fine at first...