Satish, Yes, we should change the default in petsc-dev (not release) for consistency.
Barry On Sat, 16 Jun 2007, Satish Balay wrote: > With petsc-2.3.3 the default behavior is changed for Vec [to give > error on -ve indices]. However this can be enabled with the following: > > VecsetOption(vec,VEC_IGNORE_NEGATIVE_INDICES) > > Perhaps we should do the same with Mat. > > Satish > > On Sat, 16 Jun 2007, Mark Adams wrote: > > > Thats nice to know that negative indices are ignored. convenient. > > > On Jun 15, 2007, at 10:13 PM, Matthew Knepley wrote: > > > > > Also, I have talked to Wolfgang many times about this. I am a firm > > > believer in eliminating the boundary values during assembly at the > > > element level. PETSc provides an easy mechanism for this. By default, > > > all negative indices in calls to VecSetValues and MatSetValues are > > > ignored. > >
