SSBP appears to be parallelized in c30b1, but the
parallelization seems not to work with an older mpi.

Has anyone else seen this?

It runs in parallel on an Intel Xeon running Linux (RH8.0)
with gcc-3.2 and lam-mpi 7.0.2, but runs sequentially
on an AMD machine running RH7.3 with slightly older c
libraries and lam-mpi 6.5.7.


Please notice the new address:

Thomas Simonson -- Department of Biology
Ecole Polytechnique; 91128 Palaiseau; France
phone: +33/1 69 33 38 81 (fax: -- 30 13)