Mathieu Tanneau

Extra gain of salt. Historically, I think the main reason we have "variable bounds" in the first place (as opposed to just adding them as linear constraints), is because simplex-based...

> Do you know a solver which could take advantage of knowing the matrix structure ? Tulip can. Actually, knowing the matrix structure is the main reason for having coded...

The easiest fix I can think of would be to remove this `+ CoinCpuTime()` when setting the time limit. Is it safe to do so? It feels like the `CoinCpuTime()`...

On my end, yes :)

I tried to dig into this, but this is beyond my C++ understanding. What I could gather is the following stack of calls: 1. `MOI.set(::HiGHS.Optimizer, ::MOI.NumberOfThreads(), ::Int)` (julia) 2. `MOI.set(::HiGHS.Optimizer,...

Historically, the handling of `Inf` was not clear in C and not always supported (Infinity was introduced in C99, possibly after some solvers were written). Fortunately, we don't have to...

After looking at this I'm wondering whether the approach I proposed in #7 may be a bit off-course. It might make more sense to organize things in a bottom-up approach:...

I'm still brainstorming, since I think this is an important design decision. I like more the idea of a consistent syntax across the code, and of building things on top...

This will leave one last thing to discuss/validate: the implementation of pre- and post-crush. Currently, this is done through `PresolveTransformation`s as follows: * Each time a reduction is applied, e.g.,...

> I don't find the distinction between `FixIndividualVariable` (or `FixVariable`) and `FixedVariable` The former is a rule: "If variable `j` has its lower and upper bound equal, eliminate it". The...