what about a gurobi-free version

[hariszaf]

It is my belief that we could support a gurobi-free version of DNNGIOR.

@danielriosgarza @Biomerene what do you think ?

[Biomerene]

Just to be clear, do you mean adding the option to use a different solver for the gap-filling? I think that should be possible, but I suspect it would require considerable rewriting of the gap-filling class.

[hariszaf]

Yes, I suggest we replace the gurobi parts with pytoptinterface and the HiGHS open-source solver.

PyOptInterface also supports gurobi but if we go for that milestone, we are then free-gurobi, meaning no license step etc.