Re: [cuss] Ratification of "Standards Action" guideline for draft-ietf-cuss-sip-uui

[Paul Kyzivat <[e-mail]>]

On 4/14/14 10:44 PM, Alan Johnston wrote:
> Paul,
>
> Keith and I have been discussing this in terms of new packages, which
> can define all new SIP semantics, but I notice in your response you are
> mainly talking about contents.  Perhaps we could have different
> requirements for these: the specification required for new packages, but
> something lower for contents.  I'm not completely sure where encodings
> would fit in this scheme.
>
> What do you think?

Well, if you want to do anything new then you need to define a new 
package. That must have a new purpose, and may have new content and/or 
encoding defs.

There is little in the draft that explains how those should be used 
other than the general meaning of the words "purpose", "content", and 
"encoding". The draft references the isdn draft as an example, but that 
isn't very useful for defining something more specific.

So one question is where to draw the line between purpose and content.
For instance, content might be "CSV", or it might be 
"CallCenter123.example.com". And purpose might be "CallCenter", or it 
might be "CallCenter123.example.com".

The one that is probably the least volatile is "encoding", since we are 
talking about encoding of byte strings. We probably won't need a whole 
lot of those. Maybe "CSV" would be a useful encoding, rather than a 
content. Or maybe TLV. Unfortunately we haven't set a path, so we are 
leaving it to others to decide.

In my mind, the point is that the combination used in a UUI header ought 
to be sufficient so that a receiver can determine if it understands what 
it has received. And I think in general the stuff shared between 
components of one application deployment won't work with some other 
application deployment.

	Thanks,
	Paul

> - Alan -
>
>
> On Mon, Apr 14, 2014 at 8:59 PM, Paul Kyzivat <[e-mail]
> <mailto:[e-mail]>> wrote:
>
>     On 4/14/14 9:36 PM, DRAGE, Keith (Keith) wrote:
>
>         I believe we would want the guidelines enforced.
>
>         I also expect that we would want to do a further review of the
>         guidelines if they are the sole basis for allowing packages in
>         or not, before we allowed a relaxation of the approval regime.
>
>
>     I think the effect of the registration rules will be that no new
>     registrations are made, and the default will be used by everyone, as
>     is currently the case. That means that everyone depends upon
>     configuration to ensure that client and server agree on usage.
>
>     If you are running a call center, and decide what information you
>     need to communicate from your IVR to your ACD, are you going to
>     publish an RFC to register the format for that? I think not. Even a
>     FCFS registration is probably too much. Yet it would still be good
>     to ensure that client and server are using the same formats.
>
>     When I raised this issue early on, I got no traction for it, so I
>     decided to let it go until people had deployment experience. Then
>     they could figure out they need something different. But now it has
>     come up again.
>
>              Thanks,
>              Paul
>
>         Regards
>
>         Keith
>
>             -----Original Message-----
>             From: cuss [mailto:[e-mail]
>             <mailto:[e-mail]>] On Behalf Of Paul Kyzivat
>             Sent: 04 April 2014 20:31
>             To: Alan Johnston
>             Cc: [e-mail] <mailto:[e-mail]>
>             Subject: Re: [cuss] Ratification of "Standards Action"
>             guideline for draft-ietf-cuss-sip-uui
>
>             On 4/4/14 12:46 PM, Alan Johnston wrote:
>
>                 Paul,
>
>                 I know you've explained this before.  But you haven't
>                 explained how
>                 the requirements in Section 5. Guidelines for UUI
>                 Packages can be
>                 enforced if there isn't any review.  Can you elaborate
>                 on this?
>
>
>             With no review these couldn't be enforced.
>             They could still remain as guidelines.
>
>                      Thanks,
>                      Paul
>
>                 - Alan -
>
>
>                 On Fri, Apr 4, 2014 at 9:32 AM, Paul Kyzivat
>                 <[e-mail] <mailto:[e-mail]>
>                 <mailto:[e-mail]
>                 <mailto:[e-mail]>>__> wrote:
>
>                       Interesting to see this come back.
>
>                       My original opinion was (and still is) that for these
>
>             to be useful,
>
>                       it must be possible for using enterprises to
>                 assign new
>
>             values for
>
>                       each distinct deployment of an application. IMO even
>
>             FCFS might be
>
>                       too high a bar for this.
>
>                       E.g., if I create a particular VXML application that
>
>             captures some
>
>                       data and communicates it to a call center agent
>
>             application via UUI,
>
>                       then the format of that data is likely to be unique.
>
>                       Lowering the bar below FCFS would require a naming
>                 scheme that
>                       guarantees uniqueness without registration.
>
>                                Thanks,
>                                Paul
>
>                       On 4/3/14 6:17 PM, Vijay K. Gurbani wrote:
>
>                           All: The IESG has sent draft-ietf-cuss-sip-uui
>                 to the WG to
>                           ratify the
>                           "Standards Action" guideline for defining UUI
>                 packages and
>                           registering
>                           new IANA elements for the parameter tables for
>
>             purpose, encoding and
>
>                           content.
>
>                           The draft authors note that the original concern
>
>             when the work
>
>                           was coming out of Dispatch was that the UUI not
>
>             become a "wildcard"
>
>                           header to be used for a wide variety of
>                 purposes.  Hence the
>                           direction
>                           toward requiring a standards track RFC.
>                   However, a
>
>             lesser standard
>
>                           such as "Specification Required" might suffice and
>
>             offer more
>
>                           flexibility for additional use cases, while not
>
>             opening up the
>
>                           process
>                           totally as would be the case for "First Come
>                 First Serve."
>
>                           The IESG will like to revisit this decision to
>
>             confirm that the WG
>
>                           decisions remains "Standards Action".
>
>                           To that end, Enrico and I will like to open up a
>
>             2-week period
>
>                           to ratify
>                           this decision to remain at "Standards Action"
>                 or to move to
>                           something
>                           other designation.
>
>                           The 2-week period ends on close of business (US
>
>             Central Time)
>
>                           April 17,
>                           2014.  Please express an opinion; if you are for
>
>             keeping status quo,
>
>                           please send a one-liner to the cuss WG mailing
>
>             list.  If you are
>
>                           of the
>                           opinion that we should relax the burden, please
>
>             state so and a short
>
>                           reason on why we should do so.
>
>                           Thank you all.
>
>                           - vijay
>
>
>                       ___________________________________________________
>                       cuss mailing list
>                 [e-mail] <mailto:[e-mail]>
>                 <mailto:[e-mail] <mailto:[e-mail]>>
>                 https://www.ietf.org/mailman/____listinfo/cuss
>                 <https://www.ietf.org/mailman/__listinfo/cuss>
>                       <https://www.ietf.org/mailman/__listinfo/cuss
>                 <https://www.ietf.org/mailman/listinfo/cuss>>
>
>
>
>             _________________________________________________
>             cuss mailing list
>             [e-mail] <mailto:[e-mail]>
>             https://www.ietf.org/mailman/__listinfo/cuss
>             <https://www.ietf.org/mailman/listinfo/cuss>
>
>
>