cord names

[Erik Ostrom <eostrom@research.att.com>]

>ok, i'd like to pretty much say, then, something along the lines of
>'the same rules apply to cord names as protocol names', is this ok
>with everyone?

the closer analogy is cord names to message names, not cord names to
protocol names.  cord names and message names are both dependent on
protocol names; protocol names are not.

btw, is this point addressed in the space:  if i send
"#$#mcp-negotiate-can protocol: foo-bar", does this imply i can
understand protocols foo-bar-baz and foo-baz-quux?  my vote is no, but
my recollection is that the spec implies hierarchy here, whereas my
vote is that the negotiate protocol treats the namespace protocol as
flat.