todd@CINCOM.UMD.EDU ("TODD AVEN") (10/31/86)
How many out there are using SMP to generate FORTRAN code? Have you noticed that the code generated is garbage, as is? For example, exponents that should be integer wind up being double precision, all constants are expressed as double precision numbers with LOTS of trailing zeros, etc. Speaking with Inference corporation leads me to believe that SMP can't currently be fixed to correct these problems (does anyone know if it can?). So... I've written a GNU Emacs Lisp function to take the code generated by SMP, and get rid of some of the obvious messiness. From Emacs, you simply say "M-X Fix-SMP", and viola! it's cleaned up remarkably! If there's anyone out there in the same boat as I, I'll be happy to share this stuff to help you out. Just send me e-mail... Cheers, Todd Aven Softwear Sweatshop Systems Engineering, Inc. (301)345-1692 ------