You might want to keep the mnemonic as they are for forwards combining and
invent some new ones for backwards-combining a la Unicode. Having both
certainly solves the problem, and it is valuable for me to know that all of
the ones in there thus far are forward-combining.
I have an equivalence table of composed character and their decomposed
equivalents. This has not been updated to include the material in the most
recent Unicode V1.0 spec (section 5.1) but I will do it once I know what
the mnemonics for the backwards-combining characters are. I think this
information would be useful to have available. I cannot recommend that it
appear in RFC-CHAR except as an Appendix (which standardizes nothing in and
of itself), but I'll send you a copy if you're interested in including it.
I also have a similar table for building PostScript mappings. Again, it has
not been updated to include the material in Unicode 1.0, but I plan to do
this at some point. I can send that to you as well if you want it.
I'm also looking forward to seeing how you decide to table the "allowed
combinations" information for things like T.61. A knotty problem, in my
opinion.
Keep up the good work!
Ned