[Widex] Draft minutes for WIDEX at IETF 66

[Dean Willis <[e-mail]>]

I've posted the first draft of minutes for WIDEX at IETF 66 at:

http://www.softarmor.com/widex/meets/ietf66/notes/minutes-widex-66.html

Text follows. Please review for accuracy -- they are a couple of points 
in there where I think I might have missed something.

------------------

Minutes for WIDEX at IETF 66
The WIDEX working group met on Monday July 10, 66 IETF, Montreal
Edited by Dean Willis based on notes by Dave Raggett and Vlad Stirbu

Topic: Agenda and Status
led by Dean Willis
Slides presented.

Meeting chair Dean Willis introduces the meeting and starts with the
Agenda bash.  He reminds people to use the microphone as we may be
being streamed to a wider audience.

Dave asks if he can talk about HTTP bindings before REX.  Dean agrees
to this.

Vlad launches into the presentation on the Widex Requirements.

It has gone through two revisions and has had no objections.  The next
step is to send it to the IESG for publication,

ACTION Dean to forward the Widex Requirements to the IESG


Topic: Widex Framework draft.
led by Vlad Stirbu
Slides presented

He summarizes the major updates:

 * which events have remote scope
 * clarifications on event listeners
 * clarifications on networked MVC model
     - real vs virtual view

There are 3 ways to indicate that events have remote scope

 * prior agreememt
 * negotiation at session setup, e.g. as a list of event names
 * Dynamic addition/removal during the session

The last led to the addition of a new Widex object called
WO.Selector.

Some UI markup languages don't include the means to indicate
at markup level for event bindings. This is why we defined
this new object.

A section has been added on processing external resources
such as images or sound. It is the responsibility of the
Widex Renderer to retrieve these.

 * references to external resources
 * synchronization with external resources

Next steps:

 * Add WO.* examples when next draft of REX is available
 * Add XML Processing parameters to the Service discovery
   section, e.g. encoding and charset
 * Are major things missing?

Poll: Can we consider adopting the draft framework as a Working
Group document?

Ted: Can't you reference the public version of REX and
update that as we move along?

Vlad: the unpublished version addresses the comments we
provided on the first public working draft.

Dean: when do you think the next draft will be published?

Dave: I will talk about this later

Vlad: Dave and I have done some implementation work and
it seems pretty effective.

Dean: there was quite a bit of discussion on the email list
and this has died down which is a good sign.

Eric: lets come back to the question of whether it is adopted
after we have heard about the status of REX.



Topic: BEEP bindings for WIDEX
led by Vlad Stirbu
Slides presented


This is aimed at resource constrained devices. There is a one
to one relationship between BEEP messages and Widex objects.

Dean: is there a potential need to place multiple Widex objects
in the same BEEP message for efficiency?

Vlad: REX already provides a means to do that

Dean: do we really need a new URL scheme?

Vlad: BEEP doesn't define its own scheme and leaves that to
application specifications.

Vlad asks if external resources should be handled in band or
separately.

Ted: don't put them into the same channel, and in many cases
it would be appropriate to use different streams on other
protocols. BEEP makes it easy to multiplex channels.

Dave Blacker and Andy Newton would be good
people to review this as they have done some recent work with
BEEP and are familiar with other BEEP applications.

Dean talks about two approaches to making documents WG documents.
One is to do it early as a sign of concensus and to solicit
review, the other is to wait until it is nearly finished.

What does the Director recommend?

Lisa: it seems that Vlad has a preference for the former

Vlad: does it need to wait until there are implementations?

No, early review is valuable, even before implementations.

We then went into a discussion of pipelining in BEEP. It seems
that BEEP supports pipelining naturally, but it is advisable
to use multiple channels to avoid a large resource blocking
a smaller resource. Using separate channels avoids this.

Dean suggests a large SVG document as an example.

Vlad: if the UI markup includes a reference to an SVG resource
then it is treated as a external resource and should be handled
as such via a separate channel.

If the resource is included in the UI markup then it can slow
down loading.

Vlad asks if the document is ready to be adopted as a WG document?

Dean asks Eric for his opinion. Response is inconclusive.

ACTION: Dean to raise an email poll on adoption of BEEP profile
as a Widex WG document



Topic: HTTP binding
led by Dave Raggett
Slides presented

Slides incude a good overview of the technique. Discusion followed.

Ted Hardie: one REX document from the server but is this true for the
renderer also?

Dave Raggett: yes a continous POST from the Renderer

TH: no need to to the same in both direction. one xml from the server
and many small requests from the renderer.

Lisa: ADs don't feel that the proposed extensions are in line with
the way HTTP was intended to be extended.

TH: Suggest we extend HTTP with mime types. We need to spend time to
bullet proof this HTTP binding. Solve the issues with HTTP's nature
before pushing forward. HTTP might not be the way forward to pass
beyond ajax. There migh also be problems with deep inspection proxies
that inspect HTTP POST.

DW: parallel with SIP on symmetric HTTP

TH: HTTP is stateless and you need to layer state over http.

REX Status

I Petterson: some interoperability problems might appear with event
packaging if one widex element makes assumptions about the other end.




Topic: Situation on W3C REX and influence on us
led by Dave raggett
Slides presented

The primary issue is that W3C work on REX appears to be stalled,
 following an IPR delcaration by Rance Telecom.

Action: Ted has filed a 3rd party IPR declaration to IETF.

Action: WG to review the IPR disclosures from FT and think about how
they affect us. Chair to raise discussion on-list.



Topic: SIP Binding
led by Dean Willis
no slides presented


There appears to be a consensus not to pursue a SIP binding at this
time, although the working group may revisit this at a future date.


Meeting closed with 15 minutes remaining.



_______________________________________________
Widex mailing list
[e-mail]
https://www1.ietf.org/mailman/listinfo/widex