Julian Hall

Thanks for the explanations. We will become steadily more professional in our project management, but we've been doing some of the things you suggest - like "Opening PRs allows to...

Indeed, it doesn't take much work with the simplex solver to tidy up such edge cases, particularly now I've got a primal simplex solver.

> I don't know what happens with the one failed test. There is an assertion failure in solveLp after running IPX only on mac os x: Very odd. I'll see...

> Any reason you can't use/build on SciPy's linprog function? I believe cvxpy is [taking this approach](https://www.google.com/url?sa=t&source=web&rct=j&url=https://github.com/cvxpy/cvxpy/issues/1721&ved=2ahUKEwivs5ufv5n5AhWoJ0QIHbbRC5YQFnoECAQQAQ&usg=AOvVaw2gSg9bXJkAxfYyrJhEKPFI) and you'll get support for Python specific build and usage issues Thanks for...

For the prototype "thin" Python wrapper `highspy` around the C++ class, `pip install highspy` works on two Ubuntu installations, and is not known to fail on any other Linux installation.

pip install highspy also works on (at least one) MacOS. Windows is WIP.

> Also, not part of this issue but likely a follow-up feature request, I'll be looking forward for the addition of `get/setHotStart` to the interface (keeping track in [stephane-caron/qpsolvers#94](https://github.com/stephane-caron/qpsolvers/issues/94)). Hot...

Thanks. I'll have a look at this, and add a unit test

I know that we'll need it if we are to have deterministic concurrent solution of the root node, but creating synthetic estimates of the work required to solve an LP...

I can't reproduce this. Runs OK with/without HIGHSINT64