NOTES on THMA14C for PC distribution kit

Below are notes prepared by Ken Trueblood for THMA14. These have been edited slightly and the distribution package includes test examples chosen for their relevance to THMA14c and to the paper which recently appeared:

1. The program is supplied as a 32-bit executable for Windows 95/98/NT/ME/2000.This version requires that the run-time library SALFLIBC.DLL is present in windows system directory, usually c:\windows\system for Windows 95/98 or c:\winnt for Windows NT/2000. If you have downloaded any other programs from this site you will already have this library. If not you can get it here.
2. The instructions are supplied in a separate file (MANUAL.TXT); they are extensive. First come instructions for the simplified ("NEW") input, and then the more detailed instructions for OLD input. There is a summary of the OLD input instructions at the start. Most options will not be needed by most users, but you will find it necessary to review the instructions for the "OLD" input in order to understand some features of the simplified input.
3. Test examples are included in the PC-distribution. The command line syntax for each job is given in the file TESTJOBS.TXT, which also gives a brief description of each job-purpose.
4. THMA14c is relatively new and many possible combinations of options have not been tested. It doubtless still contains at least minor problems. Let me (E.M.) know if you would like to be kept informed when we discover these.
   THMA14c should be satisfactory for most applications, but this version still does not do several things properly. In particular, when the molecule has symmetry higher than 1-bar and an attached rigid group (ARG) is in a general position, the values for the motion of that group may not always be correct. All two-atom ARGs should lead to a singular, or nearly singular, matrix when correlations of internal and overall motion are included (the default with NEW input; for OLD input, ICORL non-negative).
5. If you want to determine the mean-square torsional amplitude for a 1-atom or 2-atom ARG ('attached rigid group'), you will find that unless the group is on a site of special symmetry, it will only be done without correlations of internal and overall motion, because a 1-atom or 2-atom ARG in a general position leads to a singularity if one tries to include those correlations.

E. F. Maverick and L. J. Farrugia (May 3rd 1999)