[petsc-dev] field split/schur complement, incompressible flow

[Jed Brown]

On Wed, 12 May 2010 15:23:11 -0400, Matthew Knepley <knepley at gmail.com> wrote:
> I guess I have no problem with this if:
> 
>   a) Its completely backwards compatible with the options-style

For YAML files, I would expect to simply parse into the usual options
format.  If we had a need for the options database to represent truly
relational data, we could flip it around (changing the underlying
storage) and turn the -options_separated_by_underscores into
hierarchical options, but that seems like overkill at the moment.

>   b) the YAML part is not going to break the build

LibYAML claims to be very portable standard C, but I sure wouldn't want
it to be a required dependency (at least not until/unless it proved
itself and we saw sufficient benefit).

> I favor the low-level interface to FieldSplit taking a name and IS.

Okay, I'll do that support, and at some point (maybe in a couple weeks)
implement optional YAML parsing --with-libyaml and --download-libyaml so
we can experiment with it.

Jed