[widgets] Draft minutes 3 June 2010 voice conf

The draft minutes from the June 3 Widgets voice conference are available 
at the following and copied below:


WG Members - if you have any comments, corrections, etc., please send 
them to the public-webapps mail list before June 17 (the next Widgets 
voice conference); otherwise these minutes will be considered Approved.

-Regards, Art Barstow


       [1] http://www.w3.org/

                                - DRAFT -

                        Widgets Voice Conference

03 Jun 2010


       [2] http://lists.w3.org/Archives/Public/public-webapps/2010AprJun/0878.html

    See also: [3]IRC log

       [3] http://www.w3.org/2010/06/03-wam-irc


           Art, Frederick, Marcos, StevenP, Robin, Arve, Josh




      * [4]Topics
          1. [5]Review and tweak agenda
          2. [6]Announcements
          3. [7]Digital Signatures for Widgets spec
          4. [8]Packaging and Configuration spec
          5. [9]view-mode Media Feature spec
          6. [10]GZip, ...
          7. [11]AOB
      * [12]Summary of Action Items

    <scribe>  ScribeNick: ArtB

    <scribe>  Scribe: Art

    <darobin>  on my way sir!

Review and tweak agenda

    AB: the draft agenda was submitted yesterday (
    78.html ). Any change requests?
    ... we will drop GZip if Arve doesn't join

      [13] http://lists.w3.org/Archives/Public/public-webapps/2010AprJun/0878.html


    AB: any short announcements?

Digital Signatures for Widgets spec

    AB: the LC comment period ended June 1 and no comments were
    submitted. As such, I think the spec is ready to be published as a
    Candidate Recommendation. Any comments?

    SP: if there are no comments, it will raise some suspicion

    AB: we published a CR last summer
    ... the LCs we published since then reflected impl feedback
    ... we also got review from XML Sec WG

    SP: ok; include that data in the Trans Req

    AB: will do
    ... proposed resolution: the group agrees to publish a Candidate
    Recommendation of the widgets Digital Signature spec
    ... any comments?
    ... any objections?

    MC: Opera supports CR

    FH: I think it supports a lot of good improvements
    ... I support it

    AB: hearing no objections, I will record a positive decision

    RESOLUTION: the group agrees to publish a Candidate Recommendation
    of the widgets Digital Signature spec

    AB: who will prepare the CR version including an updated SotD?
    Perhaps we should use the WARP CR as a template for the SotD

      [14] http://www.w3.org/TR/widgets-access/)

    MC: I can do it but not until next week

    FH: when do you expect to publish?

    AB: probably not until June 22 or 24

    <scribe>  ACTION: macros notify Art when the DigSig CR SotD is
    updated [recorded in

    <trackbot>  Sorry, couldn't find user - macros

    <scribe>  ACTION: marcos notify Art when the DigSig CR SotD is
    updated [recorded in

    <trackbot>  Created ACTION-559 - Notify Art when the DigSig CR SotD
    is updated [on Marcos Caceres - due 2010-06-10].

    AB: what is the date of the earliest PR? I'd say pub date + 4 weeks

    MC: ok with me

    AB: re the pub date, how about June 24?

    SP: we want to do trans call for VMMF at same time?

    AB: yes, that is correct
    ... anything else on DigSig?

Packaging and Configuration spec

    AB: the agenda (
    78.html ) includes pointers to comments from the I18N WG. They were
    are marked as "Editorial". What is the status Marcos?

      [17] http://lists.w3.org/Archives/Public/public-webapps/2010AprJun/0878.html

    MC: I think I addressed them all

    AB: please check and make any editorial changes that are needed
    ... there was also an email from Addison Phillips the Chair of the
    I18N WG (
    63.html ). In this e-mail he voiced support for the spec changes
    Marcos has made. As such I think we have "closed the loop" with the
    I18N WG and the spec is ready for a new publication which is a
    Proposed Recommendation.
    ... any comments on publishing P&C as a Proposed Rec?
    ... Marcos, we have implementation data?

      [18] http://lists.w3.org/Archives/Public/public-webapps/2010AprJun/0863.html

    MC: yes
    ... and we can also have some implementation data for the I18N stuff

    AB: are we going to need a 2nd impl for the I18N features?

    MC: if we can show a JS impl and an Opera impl
    ... that should be sufficient

    SP: if a May, then yes, 1 impl should be enough; 2 would of course
    be better

    AB: proposed resolution: the group agrees to publish a Proposed
    Recommendation of the Widget Packaging and Configuration spec
    ... any objections?
    ... any support you want to indicate?

    <darobin>  +1

    SP: yes, go for it

    RB: support

    MC: support

    AB: I also support this

    RESOLUTION: the group agrees to publish a Proposed Recommendation of
    the Widget Packaging and Configuration spec

    AB: Marcos, please prepare the doc for publication. You may want to
    look at other PRs in /TR/ e.g. CSS3 Selectors (
    [19]http://www.w3.org/TR/2009/PR-css3-selectors-20091215/ )
    ... this will require a Director's call as well as some additional
    process e.g. AC review.
    ... I need to read up on the Process part

      [19] http://www.w3.org/TR/2009/PR-css3-selectors-20091215/

    SP: need a transition call
    ... then a vote form gets sent to the AC
    ... Must make sure that AC reps submit their vote
    ... We want to get as many votes as we can

    AB: excellent advice

    SP: do we want to include 3 specs in one trans call?

    AB: the advice I got from PLH is to keep them to 1 hour

    SP: let's first take care of the TransReq
    ... and then the call

    AB: OK; will do

    <scribe>  ACTION: barstow submit a TransReq for P&C PR [recorded in

    <trackbot>  Created ACTION-560 - Submit a TransReq for P&C PR [on
    Arthur Barstow - due 2010-06-10].

    AB: anything else on P&C for today?

view-mode Media Feature spec

    AB: last week we agreed to publish a CR of the VMMF spec
    ... Jim Allan from WAI's User Agent Guidelines WG submitted an
    e-mail (
    58.html ) about the VMMF spec. Marcos, Robin and I all responded. I
    haven't seen a reply from Jim nor the UA WG.
    ... I did ask Jim to please follow-up
    ... I am tempted to move ahead with the TransReq to CR

      [21] http://lists.w3.org/Archives/Public/public-webapps/2010AprJun/0858.html

    RB: I agree

    AB: any concerns about moving forward?

    <darobin>  "Please consider including a statement such as "The user
    agent *must* display the view-modes in a manner that meets the
    accessibility guidelines of UAAG20. ""

    AB: hearing no concerns, I will proceed with the TransReq

    RB: NB the "please CONSIDER" part
    ... we did consider it

    AB: the Plan of Record is to move forward

GZip, ...

    AB: during the last call we began to discuss GZip, streaming and
    widget packaging, etc. (
    [22]http://www.w3.org/2010/05/27-wam-minutes.html#item06 ) but Arve
    wasn't available.
    ... we can discuss this today

      [22] http://www.w3.org/2010/05/27-wam-minutes.html#item06

    Arve: the major diff between Zip and GZip Tarball
    ... is the Zip has an index at the end of the file
    ... need to wait for the zip to get loaded
    ... with GZip, there is no such index
    ... data stored in chunks
    ... header contains the data needed
    ... For packaged resources, GZip would allow immediate processing
    ... so don't have to read everything before starting to process
    ... Could place config.xml at the BoFile and then process it
    ... could process config.xml while the rest of the zip is still
    ... some widgets could benefit from this
    ... e.g. large video or audio files in the package
    ... could initialize and start game without the entire resource
    being available
    ... this is a good advantage
    ... Inserting this support into the current spec would just bloat
    the spec
    ... and delay P&C
    ... If we are to take this on, we should separate config into one
    spec and packaging into a separate spec
    ... It would then allow Tar + GZ to be in a separate packaging spec

    MC: if the market wants another signing format, we can specify one
    ... XML Sig does work
    ... we understand JAR signing could work too
    ... on a technical level, XML Sig is OK

    JS: one requirement is that it be easy to do
    ... thus our use of Zip
    ... relying on features on that are not readily available for on
    multiple platforms is not good
    ... e.g. ordering of files in a Zip varies

    Arve: if we have good use cases, tools will follow/exist

    JS: the claim that Zip can't be used for streaming I question

    Arve: but end up doing more requests

    JS: I think we have met our original reqs
    ... there is no req for partial archives
    ... I think the entire archive must be validated
    ... concerned about partial archive validation
    ... e.g. some file being deleted during the streaming
    ... that would invalidate the signed archive

    <timeless_mbp>  OK

    <timeless_mbp>  The Game use case

    <timeless_mbp>  The game has a start video

    <timeless_mbp>  which is somehow "streamable" (it sounds like in
    order to make this work it needs to be interlaced, and I suspect
    that's either split across files or not done w/ tar)

    <darobin>  [The Streaming Widget Use Case: you want to embed a widget
    in a web page. You want that to be fast. End of UC]

    <timeless_mbp>  the game also has a file which it uses to verify that
    the game is licensed to this specific user

    <timeless_mbp>  the game archive has a signature which ensures that
    the archive isn't tampered with

    <timeless_mbp>  if the archive is retransmitted and someone deletes
    that file which was used to verify the license

    <timeless_mbp>  then the author is surprised

    <timeless_mbp>  because the author was relying on the signing of the
    complete package

    <timeless_mbp>  and the package validation to protect the archive

    <timeless_mbp>  --

    AB: does anyone plan to push this into WebApps charter?

    Arve: not sure it is important enough at the moment
    ... we do need to think about market forces
    ... If there is going to be a round 2 of widget specs, we should
    consider UCs like tar-gz
    ... Should consider the spec split regardless of whether the tar-gz
    UC will be addressed

    JS: I am not opposed to making things extensible

    s/opposed as/opposed to/

    Arve: re Robin and streaming embedded widgets, agree you want that
    to be fast

    RB: don't think range requests will work

    JS: not sure tar allows interleaving

    Arve: yes, tar is one at a time

    JS: perhaps MPEG tech could be used

    RB: we need to be careful with MPEG because of W3C Patent Policy
    ... think WebM support carouselling (sp?)

    Arve: WebM is video container format
    ... not sure it is relevant for packaging web apps

    <darobin>  [and I meant interleaving more than carouselling actually]

    AB: want to stop this discussion for today
    ... but we can resume June 17

    Arve: want to propose a resolution ...
    ... to repackage P&C into packaging and config spec

    JS: I'm OK but not sure if our charter permits it

    <Steven>  I think that the *content* is chartered, so splitting
    should be trivial, since there is good reason

    AB: so is the proposal, after P&C PR is published, you want to split
    P&C into two separate specs?

    Arve: yes

    RB: I don't want to delay REC

    Arve: what about WDs?

    MC: I need to evaluate the spec
    ... it could be viewed as Editorial
    ... there could be some different ways to address the issue
    ... e.g. make it clear a different packaing format could be used

    AB: I am reluctant to record a resolution now
    ... I need to think about it

    MC: yes, we need to think this through first

    <darobin>  [I would be happy with a resolution to do that right after
    Rec, for a no-change 1.1]

    SP: if one could argue there would be no change in technical
    content, it would be safe

    <darobin>  [if we have guarantees from W3M that we can safely split,
    then fine]

    Arve: so safe to split the spec into two?

    SP: yes

    <darobin>  [additional consideration: if we split well enough, it's
    not widget-specific anymore]

    AB: I agree that if we split the spec, it would not require a
    charter review

    <darobin>  [we have Simple Web Packaging, Widget Configuration,
    Widget: Media Type and File Extension]

    AB: it could be this split would be a natural outcome of the Widget
    Embedding deliverable that has been proposed

    <darobin>  [then we just add Streamable Web Packagin]

    <arve>  [streaming web packages may very well be a core requirement
    of embedded widgets]

    <darobin>  [I think it ought to be]

    <darobin>  [without it, loading files directly will feel faster than
    the compressed, packaged version :)]


    AB: does anyone have anything to discuss?

    <darobin>  [23]http://www.w3.org/2010/api-privacy-ws/

      [23] http://www.w3.org/2010/api-privacy-ws/

    AB: no call on June 10; next call is June 17

    RB: don't forget about the Privacy Workshop!

    JS: where?

    RB: London
    ... mid July
    ... before DAP f2f meeting
    ... it is open to the Public

    Arve: one must submit a Position Paper to attend

    JS: how long?

    RB: length isn't important - cogent ideas are

    AB: meeting adjourned

    RSSAgent, make minutes

Summary of Action Items

    [NEW] ACTION: barstow submit a TransReq for P&C PR [recorded in
    [NEW] ACTION: macros notify Art when the DigSig CR SotD is updated
    [recorded in
    [NEW] ACTION: marcos notify Art when the DigSig CR SotD is updated
    [recorded in

    [End of minutes]

Received on Thursday, 3 June 2010 14:22:08 UTC