On Sat, Mar 17, 2012 at 16:00, Barry Smith <bsmith at mcs.anl.gov> wrote:
> Note that Jed's tests (below) do not catch enum values and various other > things that should also be fixed and will change the API. Yes, but enums (and other types) are not symbols, so an application with conflict only has to deal with it for the source files that include the relevant PETSc headers. It's a conflict that should be fixed, but it doesn't prevent that application from being used with PETSc in the way that symbols with extern linkage do. -------------- next part -------------- An HTML attachment was scrubbed... URL: <http://lists.mcs.anl.gov/pipermail/petsc-dev/attachments/20120317/ec22c423/attachment.html>
