Andrej Prsa

In most circumstances, but not generally. So we do need to pack all metallicities and interpolate across them.

Or perhaps to allow `run_compute()` to pass these parameters to `list_online_passbands()`? I'd just expect there to be a way to suppress the warnings because they will be inconsequential _almost_ all...

Absolutely! Would you be interested in working with us to get it implemented optimally?

Indeed, meshes in legacy are computed as part of the actual observable (lc, rv, lp, ...) and not by itself, so this is not a native operation that we can...

RVs are not less expensive, at least not by any appreciable amount. Legacy builds the entire model every time under the hood so LC vs RV makes little difference. On...

In light of github's MD support, I'd like to bump this suggestion. Pre-processing can of course always be used as a workaround, but I agree with @nilsvu that support for...

Confirming the same failure on macos-12.

@westobin, is this still an issue with 2.4.12? Can you tell us which version of c++ and python you have installed on your cluster?

@westobin, it looks like this is the same issue: https://stackoverflow.com/questions/71331756/zst28-throw-bad-array-new-lengthv-trying-to-import-c-code-in-python-after-co It seems that c++ used for compiling and for linking are not the same. Perhaps try the workaround proposed in...

Are _any_ calls to run_compute() segfaulting or just that particular one? I just tried to reproduce but it's working fine on my end.