Re: [Bug-glpk] [Fwd: Some unsuccessful runs and minor typos]

[Andrew Makhorin]

Hi Anton,

(Sorry for a very long delay in the response.)

Thank you for your bug report and the examples provided.

> The attached file contains the problem instances
> and corresponding logs indicating errors occured.

Due to improvements in the primal simplex solver glpsol 4.56 at least
doesn't crash on your instances (since the dual simplex used by the mip
solver switches to the primal simplex in case of numeric instability).
On the other hand, your mip instances are hard for glpk. In particular,
some of your instances have general integer variables in which case the
cutting planes implemented in glpk are inefficient or cannot be applied.

Best,

Andrew Makhorin