On Tue, Mar 02, 2010, Jason Swails wrote:
> Perhaps we should switch back to mpif77?
Let's leave it like we have it now. No one has reported actual problems
with mpif90, just potential ones. But we know of real problems with mpif77:
1. Lachele won't be the only one with an MPI installation where mpif77 points
to an f77 compiler not an f90 one.
2. We certainly minimize confusion by using mpif90. The fact that Amber
does not use all the possible capabilities of the mpif90 interface is of
little interest to users.
3. If someone *does* report problems caused by using mpif90, we can ask
them to see what happens with mpif77.
Let's leave it like we have it now. No one has reported actual problems
with mpif90, just potential ones. But we know of real problems with mpif77:
1. Lachele won't be the only one with an MPI installation where mpif77 points
to an f77 compiler not an f90 one.
2. We certainly minimize confusion by using mpif90. The fact that Amber
does not use all the possible capabilities of the mpif90 interface is of
little interest to users.
3. If someone *does* report problems caused by using mpif90, we can ask
them to see what happens with mpif77.
...dac
_______________________________________________
AMBER-Developers mailing list
AMBER-Developers.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber-developers
Received on Tue Mar 02 2010 - 13:00:03 PST
