JuliaIntervals / IntervalConstraintProgramming.jl

Calculate rigorously the feasible region for a set of real-valued inequalities with Julia
Other
63 stars 16 forks source link

CompatHelper: bump compat for ModelingToolkit to 7, (keep existing compat) #205

Closed github-actions[bot] closed 5 months ago

github-actions[bot] commented 2 years ago

This pull request changes the compat entry for the ModelingToolkit package from 3 to 3, 7. This keeps the compat entries for earlier versions.

Note: I have not tested your package with this new compat entry. It is your responsibility to make sure that your package tests pass before you merge this pull request.

schillic commented 2 years ago

Do you mind trying if this can be merged? To run CI automatically with CompatHelper, you need to give access to an SSH deploy key. For example, you can use the DOCUMENTER_KEY if you use it by adding COMPATHELPER_PRIV: ${{ secrets.DOCUMENTER_KEY }} here.

lucaferranti commented 2 years ago

I can try it, but I'm 99% sure that it doesnt work. Making it work with higher versions of MTK would basically be what David is doing in the rewrite branch.

lucaferranti commented 2 years ago

is it possible to use secrets.GITHUB_TOKEN instead? There seem to be no deploy key atm

schillic commented 2 years ago

I do not know the answer. Unfortunately, testing it is difficult because you need to wait until the next trigger.

schillic commented 2 years ago

By the way, I tried to run the tests locally and you were right: MTK v4 is already not supported anymore. So I guess you can close this PR (and the v6 one).

lucaferranti commented 2 years ago

yep that was my understanding too. This should be fixed once the rewrite branch is merged, but it's still a WIP. My impression was that if I close the PR without updating the compatibility, then the bot opens a new one the next day, so I kept this open to avoid spamming. Do you know how I can tell the bot "I know this doesnt work with the latest version of X, stop reminding me about that every day?"

schillic commented 2 years ago

I just checked and they have open issues about that (#257 and others), so the current behavior seems to be what you described.

dpsanders commented 2 years ago

Yes, I suggest using the rewrite branch if you need compatibility with latest versions of MTK etc.

The reason that that branch hasn't been merged is that tests are failing and some functionality is missing.

agerlach commented 2 years ago

@dpsanders can you provide any updates on the status of this? Thanks.

dpsanders commented 2 years ago

There's a rewrite branch that solves this issue. Maybe you can try that. We can probably merge that into master. Cc @lucaferranti

lucaferranti commented 2 years ago

Last time I checked (but it was a while back and should prob double check) there was an issue with ReversePropagation and teh latest version of Symbolics.jl, I have this PR based on the discussion in the corresponding issue, but I'm not sure it's right @dpsanders could you have a look at?

agerlach commented 2 years ago

OK. I was curious on the state of rewrite given that it seemed stale since you wrote

The reason that that branch hasn't been merged is that tests are failing and some functionality is missing.

schillic commented 5 months ago

This can be closed because ModelingToolkit is now optional and not in the main Project.toml anymore.