Re: [sipcore] WG adoption Request

Dale illustrates the basic mistake we’re making here, which is once again conflating the protocol state machine with the state machine of the application using the protocol. There’s no good reason to add a new response code that would semi-break existing SIP stacks just to convey a failure reason that, historically, we have other mechanisms for dealing with (like a 480 response with a Reason code).

But hey, it wouldn’t be the first cross-layer design error we’ve made in this particular cheese maze.

—
Dean

> On Oct 3, 2023, at 11:09 AM, worley@ariadne.com wrote:
> 
>> https://datatracker.ietf.org/doc/draft-khatri-sipcore-call-transfer-fail-response/
> 
> One point of concern is that this proposal needs to be useful within SIP
> as it is deployed.  Given the discussion I've seen so far, it's not
> clear operators (either PSTN or simpler SIP-over-Internet) feel
> sufficient need for this functionality to deploy it.
> 
> I also have serious questions about the detailed failure codes.  Why,
> after all, should "video call" or "real time text call" be causes of
> failure?  SIP handles all media types symmetrically and yet the current
> text wants to categorize calls according to the media they handle.  This
> is especially relevant because a call transfer always in media
> renegotiation, and so the active media after the transfer might be
> different than the active media before the transfer.
> 
> (Thinking more about this, I can imagine the fuss that deaf/real-time
> text advocate Iljitsch van Beijnum would kick up:  You can transfer
> voice calls but transfers of RTT calls are rejected!  (It's actually
> conceivable that being unable to transfer RTT calls would be classified
> as a violation of US anti-discrimination law.))
> 
> I also object to the current use of "granular" to mean "detailed" or
> "fine-grained", but that's less important.
> 
> Dale
> 
> _______________________________________________
> sipcore mailing list
> sipcore@ietf.org
> https://www.ietf.org/mailman/listinfo/sipcore