[netmod] Quirky YANG

[tom petch <[e-mail]>]

I wonder if anyone else on this list has looked at
 https://datatracker.ietf.org/doc/draft-ietf-softwire-yang/
currently in IETF Last Call.

It provides management for three tunnel types, Lw4o6, MAP-E and MAP-T
betweeen CPE and Border Relay.  It defines a CPE module, a BR module and
a module of common definitions.

It does not provide identities for the three tunnel types; rather, where
it augments "/if:interfaces/if:interface" it specifies
    when "if:type = 'ianaift:tunnel'";
which to me says that a MAP-E (e.g.) specific object will get added to
each and every tunnel of whatever type, which seems excessive.

Second, it defines four features, softwire-ce:algorithm (i.e.
MAP-E/MAP-T) and softwire-ce:binding (i.e. Lw4o6) in the CPE module and
then again softwire-br:algorithm and softwire-br:binding in the BR
module.  It is unclear to me, from the I-D, whether support is for MAP-T
and MAP-E; or either MAP-T or MAP-E - I find the I-D ambiguous on that
point - but I would have expected there to be either two - Lw4o6 and
MAP-ET - or three - Lw4o6, MAP-E and MAP-T features, not four; and for
the features to be defined, along with tunnel identities, in the common
module.

Odd, IMHO.

Tom Petch