Re: non-official updated version of glpk (4.62 pre-release)

Hi Andrew,

> i) if the primal simplex detects primal infeasibility, it should return
> rather than continue the search starting from phase I;

This should work OK as it is now.

> and ii) in the dual simplex perturbation should be disabled.

I don't understand the reason for this.


Best Regards,

Chris Matrakidis