[cuss] Final minutes for Oct-25-2011 virtual interim

["Vijay K. Gurbani" <vkg at bell-labs.com>]

CUSS WG Interim Meeting IV, Tue Oct 25. , 2011
12 participants on WebEx.
Notes taken by Enrico Marocco and Laura Liess (thanks!)

Administrivia:
Slides: http://www.standardstrack.com/ietf/cuss/cuss-interim-4/Agenda.pdf

- Agenda bashed, no changes.
- CUSS req draft with IETF, and will be sent to IESG after the Taiwan
  IETF.
- ISDN UUI Service Application Usage got matched to a draft (Keith
  Drage's draft).  A WG -00 has been submitted.


Meeting summary:

   There is no f2f meeting in Taipei, based on the agreement
   reached with respect to getting new versions of the
   mechanism and ISDN package draft out by Monday, Oct-31.
   Mechanism draft needs outside review, but ISDN package
   will need internal CUSS WG review only.  Once the new
   revisions are out, the CUSS chairs will explicitly ping
   folks for a review of the drafts. Pursuant to these reviews,
   we may need another virtual interim meeting and then
   WGLC for both of these drafts.


Alan Johnston, Mechanism draft
I-D: http://tools.ietf.org/html/draft-ietf-cuss-sip-uui-03
Slides: http://www.standardstrack.com/ietf/cuss/cuss-interim-4/cuss-mech.pdf

- Editorial changes to draft (see slide 2-6).  The changes were seen as

beneficial by some WG attendees (Paul K.). Previous comments by WG 
(Keith D.)
  on source identity of UUI have been incorporated (see slide 5), comments
  by Paul K. have also been resolved (slide 7).
- Keith asked whether we could have two different packages that use the
  same syntax with different semantics?  Discussion ensued, and it was
  decided that the draft may need more exploration on the differences
  between packages and content.  The package definition should call out
  the valid use of the content, so the discussion on syntax versus
  semantics should be put somewhere in the draft.  Alan agreed to do so.
- Discussion ensued on hex encoding (slide 8) and whether other encoding
  schemes should be provided (like string encoding).  In the end, it was
  decided to levae it at hex encoding and not worry about other schemes
  for now and wait until a package comes along that may require other
  encoding schemes.

Next steps are to close hex encoding issues, syntax versus semantics
discussion, and review by a H-I info expert before going WGLC.
Alan to get a new version out before Taipei I-D deadline.  Most issues
raised are clerical and editorial and may need f2f meeting time in Taipei.

Keith Drage, Interworking ISDN Call Control User Information with SIP
I-D: http://tools.ietf.org/html/draft-ietf-cuss-sip-uui-isdn-00
Slides: http://www.standardstrack.com/ietf/cuss/cuss-interim-4/isdn.pdf

- Went through history of the draft until the adoption in WG.
- Outlined issues (slides 4-7); none contested or needed any long
  discussion time by WG attendees.
- Slide 8 (Issue 5) resulted in some discussion time regarding removal
  of diversion mechanism and whether the length limitation pertains to
  the package or the entire message.
- Some discussion on what to do when the header field appears
  multiple times for the same package.  One interpretation is that you
  can pick any of them.  Other could mean that you pick all of the ones
  you support.  Some choices need to be made on what sematics to apply
  to multiple headers.  James reminded that at the last IETF meeting
  we decided not to worry about multiple headers due to this very
  complexity.  However, at this point the ISDN UUI package is the only
  one, so it is hard to generalize based on one package.
  Alan will take a first cut to put some text in the mechanism draft
  on how to handle multiple headers and separate package syntax from
  semantics.  The discussion belongs there and not the ISDN UUI draft.
- Keith believes that pursuant to a full review by CUSS WG, the draft
  is ready for WGLC.
-------------------------------------------------------------------

Raw notes from Enrico Marocco

CUSS WG Virtual Interim Meeting
October 25, 2011

Chairs: Vijay Gurbani <vkg at bell-labs.com>
	Enrico Marocco <enrico.marocco at telecomitalia.it>

Start time: 17:05 CEST


Agenda
------

Vijay presents the agenda for today and the status of the WG.


Keith: I expect the Taipei meeting to be confirmed today. Vijay: we'll 
decide at the end of the meeting if we need to meet in TW.
Mechanism draft
---------------


Alan Johnston presents version -03 of the mechanism draft. Terminology 
changes to make the use of "UUI data" and "package" consistent through 
the document
  Paul Kyzivat: I like it!


Updated ABNF. Clarified that source identity is source or request or 
response, in From, To, P-A-ID.
Keith Drage: this is not what we discussed in the past. I'd like to see 
text saying that the source of UUI is the source of the SIP message 
(either req or res). Alan: that is the intent of the text, read it and 
suggest changes if it does not sound like that to you. Keith: will do.
Added the IANA considerations section. Open issue is about how packages 
define content types.
Keith: could we have two different packages that use the same syntax 
with different semantics? Alan: content is about syntax, package is 
about semantics. Paul: may there be a case where multiple content types 
may be needed? Wondering whether there's value in having flexibility on 
that. Keith: most people usually use one content type only. Alan: when 
discussed in DISPATCH, people had concerns about passing blobs of data 
without saying what's in it. That's the case with ISDN. Do you think the 
draft needs more investigation about the difference between content and 
package? Paul: yes, the discussion about syntax and semantics is 
probably worth putting somewhere in the doc.
Proposal to include text proposed by Paul to specify the hex encoding.


Paul: if we go this far, should we also define other encodings? A string 
encoding? Alan: basically everyone uses hex today. James Rafferty: we've 
seen both text and hex, but nothing prevents converting a string to hex. 
Paul: if a conversion is needed in the ISDN, it can be done in the 
gateway. Keith: prefer to keep the gw simple, just have it copy UUI from 
ISDN to SIP and vice versa. Paul: there may be value in defining a 
string encoding, but don't feel strongly about that.
Will add in the next version two things: syntax vs. semantics and 
details of hex encodings. Next review and hopefully WGLC.

ISDN usage draft
----------------


Keith presents the freshly adopted version -00 of the ISDN usage draft. 
Terminology changes to align the ISDN usage draft to the mechanism one, 
plus some issues suggested on the list.
Alan: is fine with removing the contributor's note at the end of section 3.

Want to check whether people think this would define a valid package.


Alan: what do you think about the removal of the diversion mechanism? 
Keith: there's no way to interworking it.
Alan: for the multiple values, the ABNF in the mechanism right now 
forbids multiple header fields. If we want to allow them, we need to 
modify the ABNF to allow combined comma-separated header fields in the 
same header field. Paul: don't think 3261 says exactly that. Alan: we 
can say that if the header field appears multiple times, they are 
basically alternatives. James: at the last IETF meeting, we decided not 
to allow multiple headers, for complexity. Alan: but multiple headers 
defining alternative representation of the same UUI may make sense. 
That's something for the mechanism draft though. Paul: the mechanism 
should provide some indication about what package definitions can and 
should define.
The document is basically ready for WG review, WGLC and so on.


Paul: based on earlier discussion on encoding, it should be defined in 
such a way that is orthogonal to the content definition. Keith: it 
currently mandates hex encoding. Paul: and what about other possible 
encodings? Keith: that would break interoperability the way we have 
thought of so far. Alan: this ISDN usage is an example of why a package 
may want to restrict the encoding. Paul: I still think there's value in 
keeping content and encoding orthogonal, will take it to the list. 
[Further discussion on the content/encoding dichotomy, about e.g. 
whether coders or call-center admin or other standards bodies will be 
the one defining content.]
Next steps
----------


Both Alan and Keith will be able to revise the drafts by the -0x 
deadline for Taipei.
Keith: it would be very useful to have assigned reviewers from both CUSS 
and DISPATCH.
Vijay: do we need to meet face to face in Taipei? Keith: for the ISDN 
document don't see a need. Alan: will we have reviews before TW? Vijay: 
unlikely. Alan: then would probably have little to discuss face to face.

End time: 18:20 CEST
---------------------------------------------------------------------

Raw notes from Laura Liess

Notes CUSS interim Oct 25, 2011


Participants: Vijay, Enrico, Alan, Keith, James, Paul, Laura (I don’t 
know who pm was in the webex participants list)
1) Vijay’s presentation
Note Well statement
Agenda. Interim participants agree to the proposed agenda.

Administrative issues. It is not quite clear if the time slot for the 
Taiwan meeting is needed.
2) Alan’s presentation on the UUI mechanism:

Recent changes (slides 2-6): The participants agree to them. Slide 5: 
Keith pointed out that the source of the UUI is the source of the 
request or response. Issues raised on the list and Paul’s proposal for 
hex encoding (slides 7-9). Discussion about the need for more 
clarification on syntax vs. semantics and hex encoding (mainly Keith, 
Paul, Alan). Conclusion of the discussions on the raised issues: The 
next version of the draft will add text for:
Syntax vs. semantics
Details of hex encodings

-  Next steps:
Alan will try to submit a new version of the draft before next Monday
No need for a f2f meeting in Taiwan identified.

3) Keith’s presentation on the ISDN package
History since last Ad-hoc (slides 2-3). No additional comments.
Issue1: Alan: The change is OK.
Issues 2 and 3: ???
Issue 4: Keith asked for suggestions. No suggestion for now.

Issue 5: Discussion if some description is needed concerning multiple 
header fields. Alan: There was text on this in the past and it was 
removed. He offered to provide the text on this issue for the next version.
Additional technical discussion:
Paul:???

Alan: The ISDN-package has constraints to keep the gateway 
implementations simple.
Next steps:
Keith will try to revise the document till Monday

No need for f2f meeting in Taiwan identified, further discussion on the 
list.
After Taiwan, try to find more reviewers, also from other WGs.



- vijay
--
Vijay K. Gurbani, Bell Laboratories, Alcatel-Lucent
1960 Lucent Lane, Rm. 9C-533, Naperville, Illinois 60566 (USA)
Email: vkg at {bell-labs.com,acm.org} / vijay.gurbani at alcatel-lucent.com
Web:   http://ect.bell-labs.com/who/vkg/