[Dots] on cyclical normative dependencies

[Benjamin Kaduk <[e-mail]>]

During the session today, Med mentioned at the mic that if we moved the
content from draft-nishizuka-dots-signal-control-filtering into the signal
channel document, then we would have a normative dependency from the signal
channel on the data channel (and the data channel already depends
normatively on the signal channel), as if this would be problematic.

I don't think this is inherentely problematic, as it just means that the
two documents will advance together as a group.  It's still a useful tool
to be able to separate the content out into logically separate documents
even if they are interdependent and in some sense must be considered
together as a consolidated group.  We see this pattern occur regularly in
the IETF, and it's not worrisome.

So I don't think that the cyclical dependency is a reason to leave
signal-control-filtering as a separate document (but I do agree with the
sentiment in the room that we may want to let this document settle in for a
while before we decide that we're happy with it), but given the trend of
the discussion it didn't seem worth spending the time to say it at the
microphone.

-Ben