On 4/27/20 4:11 PM, Brian Rosen wrote:
Telephone numbers are actually very interesting, useful, and resilient
identifiers. Try taking your email address to a different provider.
Or your sip URI.
Try verbalizing your sip URI so you can give it to someone without
first getting something from them (and how did you get that?).
Unlikely that SIP URIs will replace telephone numbers in my lifetime.
This doesn't seem to be a problem anywhere else. My devices hide all of
that from me these days. Seems like a very marginal benefit.
The stir part of STIR/SHAKEN works fine on sip uris. The SHAKEN part
doesn’t, but could be extended to pretty easily. I’m sorry you found
it hard to understand. We have a heck of a lot of deployment
underway, and the dev teams don’t seem to need a lot of help, so the
documents are fairly good. I’m not close to the implementation work
any more, but I think interoperability continues to be excellent.
We had interoperablity before we even had a working group for DKIM.
Iirc, we had multiple interoperable versions before the working group,
but then again it took forever to get the wg spun up.
I am heavily involved in the NANC IVC interoperable video conferencing
work. It’s very interesting, but right now it’s voluntary. Many of
the big players are participating, but there are zero commitments to
deploy as yet. We’re heading towards using new routing protocols (so
Shockey doesn’t need to be “convinced” to resurrect enum). Probably
something along the lines of modern. No one wants a new monopoly
player. There will be new IETF work coming out of IVC I think. It
has features to support all sorts of assistive devices that need some
form of interpreter, but NANC won’t do any signaling work — they are
the telephone number routing experts.
I'm reading rfc 8824 and I am aghast at the number of things that are
just flat out wrong. I haven't compared notes with Crocker's last call
comments, but I have no doubt that we share a lot of the same concerns.
Which, manifestly, were ignored since they are still there. Some of them
are actually pretty dangerous because they make claims that cannot work
in reality, and where the confusion can be can be exploited.
Monopolies are a layer 8 problem. The only question is whether SIP
conferencing will be used in the same way that email is used going
forward completely ignoring all of the legacy bellheaded stuff. If this
is SIP's answer to DKIM, then it missed a lot of DKIM's lessons learned.
Mike
PS: I was rather amused to see the word Bellhead used in an article in
the Economist a couple of weeks again.
