[Clp] CLP finds feasible model infeasible

[Andrew Makhorin]

> This time, I forgot to mention it. It's a maximization problem

> Sure you can have a free variable with a non zero objective.
> 
> This model has a primal dual optimal solution :
> 

>> Besides,
>> since that column has a non-zero objective coefficient, your model is
>> dual infeasible.

Oops, sorry. In fact, the column 'futureValue' has an implied bound,
which doesn't appear to be active in case of minimization. The model
infeasible_65899_150.mps is primal feasible, though badly scaled, and
has a finite maximum OBJROW = 891245.5383 (I checked that with the glpk
simplex solver).


Andrew Makhorin