[molpro-user] {SPAM}? mcscf optimisation

[Alister Page]

Hi,
Im using Molpro 2002.3 and am having some problems with the optg routine.  im trying to optimise Be2 at the mcscf level.  my input deck goes like,
***,
memory specs
geometry stuff
basis stuff
hf;wf,8,1,0;
save,2100.2;
multi;
core,1,0,0,0,1;closed,1,0,0,0,1;
occ,5,3,2,0,6,1,1;
freeze,3.2,6.5,1.6,1.7;
wf,8,1,0;start,2100.2;
optg;
---

molpro comes back with 

Sorry, MCSCF gradients with frozen core not yet implemented in Alaska.  Please try to use gradtyp=cadpac

Trying this, instead of optg i gave

optg;forces,gradtyp=cadpac;

and Molpro comes back with the same error....  Trying numerical gradients instead of analytical gradients with

optg,numerical,startcmd=hf;

Molpro comes back with 

?START RECORD 2140.2 CORRESPONDS TO DIFFERENT GEOMETRY

in the .log file (which makes no sense to me, shouldnt multi read from 2100.2?)

Can anyone help me get around this problem?
thanks in advance!

Alister Page
Molecular Structure and Detection Group
School of Environmental & Life Sciences
The Faculty of Science & IT
The University of Newcastle
University Drive, Callaghan, NSW, 2308.
Australia
Room: LS4.03 (Life Sciences Building)
Ph:  4921 7894
Homepage:  http://msdg.science.uncle.newcastle.edu.au/alister/index.htm
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.molpro.net/pipermail/molpro-user/attachments/20050701/b0e797e4/attachment.html>