> > It's a normal int, not a PetscInt, so at least it will always fit in that > pointer variable. > > You can silence the warning by changing the line from > > leftarray[0] = (point) i; // The array length. > > to > > leftarray[0] = (point) (PETSC_UINTPTR_T)i; // The array length. >
What I don't understand, is that with all this work done (thanks Matt!) why is it bundled with PETSc and not as a forked project? I mean, ctetgen seems like it would have value for other projects and most of the work is just scrubbing out the C++ (which, again, Matt just did the bulk of!). Why not just make an interface to it? This would have the added bonus of solving the license quirks (and, yes, Matt, there is a difference between having a fork / linking vs. *bundling* the code). -------------- next part -------------- An HTML attachment was scrubbed... URL: <http://lists.mcs.anl.gov/pipermail/petsc-dev/attachments/20120319/e7dbe146/attachment.html>
