Benjamin Lorenz

@fingolfin please try merging my wip commit from here: https://github.com/benlorenz/Yggdrasil/commit/502048c6985a4a64e9dfacda207fa3fb21c8fd43

As far as I understand, this does not grant a company like Julia Computing a license to redistribute it (or use it), also the installations are probably not tracked according...

> If you write a package that depends on SCIP, you should do due diligence on the license restrictions. It also doesn't install automatically on Windows, which likely limits the...

Thanks! I have adjusted the first post and the title, I would like to keep this open for the boost and bliss issues, but I can have a look at...

@lgoettgens I am not really sure why this should be in 1.0, it is mostly meant to help keep the code-bases separate but there will not be much further development...

Maybe it would make sense to wait with merging this PR until we are done with most of the backporting, to avoid any conflicts, both for master and for the...

I changed it to draft so that we don't accidentally merge this right now, we can change it back some time next week I guess.

I have rebased this branch on the current master, implemented the changes suggested by Dan and moved the initialization of the RNG to the top `runtests.jl` file.

The timings look significantly better than on the current master, even nightly on macos finished (but that one might just be a lucky coincidence).

I marked is as a draft for now as I need to change where the `rng` initialization happens as the current approach break `test_module(...)` and including a test file directly.