[molpro-user] Re: Re: molpro on Opteron with SuSe

[Andrew Ichimura]

Molpro 2002.6 was recently complied (i8) on my new Opteron system with 
SuSe.  All test jobs ran to completion without error.  Now, the program 
is behaving oddly.  An input file that ran yesterday fails to complete 
today, the only difference being the print options, but the problem 
started a week ago.

What exactly are the "unresolved issues" that make the i8 build "less 
stable" than the i4 build?

This following has been driving me crazy:

*******
Geometry input for large molecule, C2v symmetry, two states 3B1 followed 
by 1A1 MCSCF
basis=vdz
rhf;
wf,140,2,2
occ,31,17,16,7
multi;
occ,140,31,17,16,7          !Essentially an ROHF calc using Multi code 
to show that it works,
closed,140,30,16,16,7      ! and it does.
multi;
wf,140,1,0             ! 1A1 singlet state at singlet geometry
occ,31,17,16,7       ! two electron/two orbital CAS;   FAILS HERE!
closed,30,16,16,7
RS2;

Molpro does not read the OCC and CLOSED cards correctly and instead 
crashes with the error
"Too many active orbitals"
Number of closed orbitals:  20 ( # of core orbs in each sym)
Number of active orbitals:  96  ( " )            
**********

This is the full valence space and clearly not indicated in the input!!!

Any information on this is greatly appreciated!

Andrew

-- 

Andrew S. Ichimura, Assistant Professor
Department of Chemistry and Biochemistry
San Francisco State University
1600 Holloway Ave.
San Francisco, CA 94132
Tel:  415-405-0721
Fax: 415-338-2384
ichimura at sfsu.edu