Oscar Dowson

Your models are not equivalent because the lower bounds are different. In JuMP, if you want a large bound, just omit it. In C++, use `highs.getInfinity()` for the value that...

Okay, so I ran this with Gurobi as comparison: ```Julia julia> using JuMP, HiGHS, Gurobi julia> function run_model(optimizer) model = Model(optimizer) @variable(model, v_0, Int) @variable(model, v_1, Int) @variable(model, v_2, Int)...

> The implied lower bounds explain this. JuMP doesn't have implied lower bounds. The variables are free with `-Inf, Inf` bounds.

Here's what I get with v1.2.1: ``` shell> cat problem.mps NAME ROWS N COST L R0 L R1 L R2 COLUMNS MARK0000 'MARKER' 'INTORG' C0 COST 1 C0 R0 1...

Do we care that we don't support i386? According to https://en.wikipedia.org/wiki/I386, Linux dropped support _10_ years ago.

I would caution against adding specific callback functions. It's a lot easier to deal with a generic callback infrastructure with codes to indicate which callback is being run. (There's a...

Should this issue be closed or renamed to something like "OSI interface needs updating"?

I haven't tried linking the binaries to C++; we only use the C interface. Did you try dynamically linking with `/bin/libhighs.dll`? If that doesn't work, HiGHS is not difficult to...

Caveat: I don't have any experience compiling HiGHS on Windows, or linking a DLL to C#. > which is not available in Windows https://stackoverflow.com/questions/32127524/how-to-install-and-use-make-in-windows > Is it possible to solve...

Please provide the MPS file and any options that were set.