I was going over some pages today from a prog lang/comp opt person, trying to figure out if any of their stuff could apply to what we do. It was very hard, because there were a lot of general methods (incrementalization, loop transformation) that are known to fail in many cases. I found myself wishing that they had made a list of algorithms, perhaps with the relevant citation, for which these methods were known to work.
It then occurred to me that we are in a very similar situation. I know its a pain in the ass because usually it requires some impossible to write preconditioner, lots of extra domain information, etc. However, just listing a few clear papers for major problem areas would be a great resource I think, and we could point a lot of petsc-maint queries that way. It might also help clear up our discussion on whether we can actually do some things, like solve variable viscosity Stokes. Matt -- What most experimenters take for granted before they begin their experiments is infinitely more interesting than any results to which their experiments lead. -- Norbert Wiener -------------- next part -------------- An HTML attachment was scrubbed... URL: <http://lists.mcs.anl.gov/pipermail/petsc-dev/attachments/20110604/b0db0e19/attachment.html>
