Dear all,
>
> If there is no strong objection expressed against this solution in the
> coming days, I will go for the first option (i.e. the pair
> getResidual()/providesResidual()), as it avoids having to define
> tagging interface which I don't think would be very useful.
>
I've now implemented
Hi again,
>
> Dear all,
> I know it's not the time to open new issues, but this one should be
> quickly solved (I promise!), and I think it would really improve the
> whole framework for iterative linear solvers. Anyway, please have a
> look to what I've written. Also, I really need some feedback
Dear all,
I know it's not the time to open new issues, but this one should be
quickly solved (I promise!), and I think it would really improve the
whole framework for iterative linear solvers. Anyway, please have a
look to what I've written. Also, I really need some feedback on point
#4, which I r
