On Mar 3, 2009, at 11:15 AM, internet-drafts@xxxxxxxx wrote:
A New Internet-Draft is available from the on-line Internet-Drafts
directories.
This draft is a work item of the Session Initiation Proposal
Investigation Working Group of the IETF.
At the risk of sounding like Henry, I think we have good cause to
worry when we consider that we are on the 12th rev of our call
transfer draft, and that the SIPPING working group version was
preceded by 5 versions in the SIP working group. Even earlier, we had
the original one individual version published in March, 2000.
What kind of monster have we created that it takes nine years and 18
versions of a draft to specify call transfer, and we're still not
done? The kudzu has eaten the crop, and it may be time to just spray
Roundup on the whole field and replant. Learning to eat kudzu is not a
viable option.
Seriously, we need to step back, take a deep breath, and think very
seriously about simplifying the whole SIP framework instead of
elaborating it further. I don't know that I completely agree with
JDR's modest proposal about accepting the SIP-as-SBC-mediated-
telephony-over-IP model as our entire scope, but we have to do
something.
How about freezing SIP 2.0 in maintenance mode, taking on no further
work (except for bug fixes), and starting a SIP 3.0 exercise that uses
things we learned from SIP 2.0, including but not limited to:
1) Isolating the application, transaction, rendezvous/naming, and
transport layers.
2) Supporting, or better, fundamentally integrating strong end-to-end
identity.
3) Recognizing and formalizing the roles of active intermediaries that
exceed the limitations of SIP 2.0 proxies.
4) Limiting optionality.
5) Including strong compliance specifications that can be tested
against and certified independently.
I'd also like to put in some text about reusing/sharing connections
and muxing signaling and media onto an address/port pair, but I'm sure
that would cause somebody around here to self-destruct. Perhaps this
could be accomplished by:
6) Assume that NAT, NAPT, and address family (IPv4/IPv6) translators
are ubiquitous, and based on this assumption optimize the protocol
design so as to minimize the protocol adaptation mechanics required to
traverse those translators. In other words, stop referring to
addresses and ports inside the payload.
Note: From a metrics perspective, it is interesting that the draft has
been revised just over twice for every three meetings, which may go
towards disproving the "drafts are revised for each meeting" theory,
or perhaps restating it as "Drafts are revised, on average, no more
than once for each meeting cycle."
--
Dean Willis, frustrated SIP gardener.
_______________________________________________
Sipping mailing list https://www.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@xxxxxxxxxxxxxxx for questions on current sip
Use sip@xxxxxxxx for new developments of core SIP
