Come on David, you can't drop bombshells without giving us the
juicy bit.

db> For example, the infamous "64-state trellis coding" - s9.6.3.2 
db> ("Convolutional Encoder" - part of the trellis encoder)
says that *three* 
db> encoders are *available*:  16-state, 32-state and 64-state.  It doesn't say 
db> that all three should be implemented.

db> I haven't gone completely through this document that I don't really have, 

Explain please!

db> but from what I have seen, there's no canonical statement about the type of 
db> trellis coding that should be supported in order to claim V.34 compliance - 
db> 'cept of course that it's gotta be one of the three above.

So that's the canonical statement?  The spec IS the spec!  What
more do they need to say?  There is a flag, the flag must have
one of three values, what's the problem?  There's probably a
statement about what to do if one person has 64 and the other
has 16.  Probably something about 16 needing to be supported
as a common minimum, or that the modems can do without it altogther
or something.  More likely, 64 means you support UP TO 64, ie
you also support 16 and 32.  BFN.  Paul.
@EOT:

---