dolfin team mailing list archive

[Matthew Knepley <knepley@xxxxxxxxxxx>]

Anders Logg <logg@xxxxxxxxx> writes:
> There are a couple of different options we could consider:
    I can comment on the ones I know about.
> 
> 1. PETSc: http://www-unix.mcs.anl.gov/petsc/petsc-2/
> 
> Pros: Satisfies B1-4
> Cons: Does not seem to satisfy B5. Maybe a future version of PETSc will?
     PETSc proper will never move away from the C interface. However, I
have already written Python wrappers with a fully OO interface (not hard).
I used SIDL, so the same method could generate C++ wrappers as well.
> 
> 2. MTL: http://www.osl.iu.edu/research/mtl/
> 
> Pros: Satisfies B1 (?) and B5
> Cons: Does not satisfy B2-4?
     I believe they are working on parallelizing it, and Andy does maintain
this, although I think it does not really have a large customer base yet. It
is definitely open source.
> 
> 3. uBLAS: http://www.boost.org/libs/numeric/ublas/doc/index.htm
> 
> Pros: Satisfies B1-3, B5
> Cons: Does not satisfy B4
     I don't think parallelism will even happen here. It is much more likely
from Andy Lumsdaine and MTL.
> 
> 4. POOMA: http://www.codesourcery.com/pooma/
> 
> Pros: ?
> Cons: ?
     I would not even consider using this.
> 
> 5. Sparselib++: http://math.nist.gov/sparselib++/
> 
> Pros: ?
> Cons: ?
     Or this.
> 
> 6. TNT: http://math.nist.gov/tnt/
> 
> Pros: ?
> Cons: ?
     Or this.
> 
> 7. Blitz++: http://www.oonumerics.org/blitz/
> 
> Pros: ?
> Cons: ?
    I like this project since it is so innovative (and Todd is cool). However,
I did not think it was ever intended to be parallel, and I am not sure
what the support for sparse matrices is like, but I think it might even be
missing.

    I would note that you have not considered Trilinos, which is a library
very similar to PETSc from SNL.

        Matt
> 
> /Anders
-- 
"Failure has a thousand explanations. Success doesn't need one" -- Sir Alec Guiness