[petsc-dev] MatNest considered harmful

[Jed Brown]

Barry Smith <bsmith at mcs.anl.gov> writes:

>   Maybe just remove MatNest :-)

Seems like it would be enough to remove the MatNest*() functions.  The
problem is that people specialize their code for MatNest, not the use of
that data structure as an optimization.  It seems like no matter how
much we warn people, they can't help but make their code brittle.  Maybe
we need to find a clearer interface/explanation of doing it the right
way.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 818 bytes
Desc: not available
URL: <http://lists.mcs.anl.gov/pipermail/petsc-dev/attachments/20140812/df818fe1/attachment.sig>