A proposal for moving ahead on BC models

[Jim Boyle <jboyle@pdnets.com>]

Francois, thanks for getting this conversation rolling.  Let's hear from 
other members of the WG.

At first, we had this question as "should there be 0, 1 or more
required bandwidth constraint models that a diffserv TE implementation
must support?".  However, since that is a complex question, we broke
it out.  Here's the structure of the question as we had it at the
meeting.

1) Should there be a specified (required) default bandwidth model
	that diffserv TE implementations must support.

2) If so, so there *only* be one?

3) and if so.... which ?	

Granted, 3 wasn't explicitely discussed.