CI Convergence problem

[Sudhakar Pamidighantam]

We observed that the CI is not converging after the latest patches were 
added for the following input (which converged smoothly before )
***, C3H3 + H -> singlet allene, Cs symmetry
text, Gaussrate reaction path, s=0. (C-H = 3.5 A), E=-0.0994 kcal
memory,25,m file,2,allene1.2
geomtyp=xyz basis=cc- pvtz
gprint,
distance,basis,angles=1
geometry= {
7
s = 0.0000 bohr V = -0.0994 kcal
C 0.08474620 0.00000000 - 1.26006439
C 0.08474620 0.00000000 - 0.03374410
C 0.08017900 0.00000000 1.36627771
H 0.10532819 0.00000000 - 2.32201733
H -3.23537428 0.00000000 - 2.36767442
H 0.07862568 0.93062180 1.91340651
H 0.07862568 -0.93062180 1.91340651
}
mcscf; occ,11,3;
closed,7,1;
start,2140.2
wf,22,1,0
accuracy,energy=1.d- 9
orbital orbprint
ci; thresh,energy=1.d- 6
  ---
Has something changed in the way CI methods are used that I missed.

Thanks,
Sudhakar.


---------------------------------------------------------------------------- 
------------------------------
Sudhakar Pamidighantam, Ph. D.      Ph:217-333-5831 Fax:217-244-2909
NCSA @ UIUC, 4147 Beckman Institute email: spamidig at ncsa.uiuc.edu
405 N Mathews Ave. Urbana, Illinois 61801 WWW: swarna.ncsa.uiuc.edu
---------------------------------------------------------------------------- 
-------------------------------