Brandon Bocklund

https://github.com/pycalphad/scheil/pull/39 - currently `simulate_scheil_solidification` works, but several things are broken. The idea in connection to this is that you could ask for things in `simulate_scheil_solidification(..., output: list[str])` and then per...

(Take my comment with a grain of salt: I've participated in the JOSS community but not in the pyOpenSci community). Part of the [Guide for Reviewers](https://www.pyopensci.org/contributing-guide/open-source-software-submissions/reviewer-guide.html) suggest that the UI...

We should verify that leading zeroes are parsed correctly by commercial software

Thanks for the reminder ping! I did also verify that this test fails for PyCalphad 0.10.5 (pre-Workspace), so this doesn't seem like a regression (which is good), but still is...

I don't think I have a database with that set of components and I'd guess the convergence failure is more sensitive to the particular Gibbs energy surface rather than the...

I think this should be resolved in the `develop` branch with #605 now merged. @CountryTree3 can you provide a database to test against or try again with your database on...

I am able to reproduce this on a Windows machine. It doesn't happen every time you run the tests (and, in my experience, less frequently than on the CI tests)...

Downgrading to symengine 0.13, script (2) above no longer produces failures. So this seems to be an upstream issue and pinning symengine to 0.13 might be a mitigation for now.

Mentioned upstream that this could be related to some memory collision in how newer versions of symengine handle serialization. One (untested) idea for a workaround is to try caching our...

The main question for now is whether to try to raise this upstream in SymEngine, try to fix locally, or to change the `__getitem__` API for getting new units for...