I have very strong doubts about all the issues mentioned in sections 4 and 5
I have comments on
4.2 TDM PWs,
4.3 Codecs, and
4.4 MPLS signaling protocols.
For TDM PWs the IETF followed its tao of requiring one MUST mode (RFC 4553)
and allowing 2 MAY modes (RFCs 5086 and 5087). However, far from being a
triumph, this meant that a method considered inadequate by all who really
understood the issues
was mandated, while too equally adequate alternatives became optional.
So this example actually shows a negative aspect of the procedure.
Regarding the Code issue, multiple proposals were submitted,
and the resolution was to mold two of them into a new combined codec with
pieces of each.
So this example is not relevant at all.
Regarding CR-LDP vs. RSVP, once again there were two pre-existing solutions,
and it was decided not to do any new work on one of them.
But this left us with 2 mandatory to implement protocols (LDP and RSVP-TE)
rather than a decision for CR-LDP which would have left a single mandatory
So this actually a counter-example to the tao of having a single mandatory mode.
Ietf mailing list