Now that we can get Mat ownership range/reanges from rows and cols, would it make sense to condensate the calls into a single MatGetOwnershipRange[s] able to return all the ranges? Something like: MatGetOwnershipRange(Mat A, PetscInt* rlow, PetscInt *rhigh, PetscInt *clow, PetscInt *chigh); MatGetOwnershipRanges(Mat A, const PetscInt* rowranges, const PetscInt *colranges); We could still provide MatGetOwnershipRange[s]Row() and MatGetOwnershipRange[s]Column() and implement the first to calls above in terms of these later four. Of course, this smells to too many ways to do it. Note that I'm +0 on this, I just wanted to point out some other choices. -- Lisandro Dalc�n --------------- Centro Internacional de M�todos Computacionales en Ingenier�a (CIMEC) Instituto de Desarrollo Tecnol�gico para la Industria Qu�mica (INTEC) Consejo Nacional de Investigaciones Cient�ficas y T�cnicas (CONICET) PTLC - G�emes 3450, (3000) Santa Fe, Argentina Tel/Fax: +54-(0)342-451.1594