W3C home > Mailing lists > Public > public-xg-htmlspeech@w3.org > December 2010

[minutes] draft minutes 2 December 2010

From: Dan Burnett <dburnett@voxeo.com>
Date: Thu, 2 Dec 2010 17:21:01 -0500
Message-Id: <5857F7BB-7C15-41C0-8A83-1BF01B52AD05@voxeo.com>
To: public-xg-htmlspeech@w3.org
are at http://www.w3.org/2010/12/02-htmlspeech-minutes.html.

A text version follows for your convenience:


           Bjorn_Bringert, Dan_Burnett, Olli_Pettay, marc,
           +1.732.507.aaaa, Milan_Young, Debbie_Dahl, Raj_Tumuluri,
           +44.207.881.aabb, Satish_Sampath, [Microsoft]

           Dan Burnett

           Raj Tumuluri


      * [4]Topics
          1. [5]minutes from last week's call
          2. [6]Updated Requirements Document
          3. [7]R14
          4. [8]R25
          5. [9]R24
          6. [10]R10
          7. [11]: R20
          8. [12]R23
          9. [13]R12
         10. [14]R23
         11. [15]R12
         12. [16]UA/SS
      * [17]Summary of Action Items

    <burn> trackbot, start telcon

    <trackbot> Date: 02 December 2010

    <smaug_> um, seems like my network connection is pretty bad today

    <bringert> I am Bjorn_Bringert

    <burn> Scribe: Raj Tumuluri

    <burn> Scribenick: Raj

minutes from last week's call

    Minutes Approved

Updated Requirements Document

    No objections to the new requirements document


      [18] http://lists.w3.org/Archives/Public/public-xg-htmlspeech/2010Dec/0015.html

    Bjorn proposed rew-wording..are there any objections?

    No Objections from the group


      [19] http://lists.w3.org/Archives/Public/public-xg-htmlspeech/2010Dec/0014.html

    There were some recommendation to the wording ....are there any

    No objections from the group


      [20] http://lists.w3.org/Archives/Public/public-xg-htmlspeech/2010Dec/0018.html

    No objections recorded for the above requirement

    Topic R2

      [21] http://lists.w3.org/Archives/Public/public-xg-htmlspeech/2010Dec/0023.html

    Proposal to replace R2 with two new requirements sent through emails
    ( last email on Monday)

    1. Recogntion w/o specifying grammars should be possible

    milan: Did we agree to specify language as part of grammar

    Bjorn's wording: Should be possible to specify the language
    completely independently of the grammar

    <burn> bjorn: we already have a requirement to address this

    <smaug_> FPR38.

    Milan: I did not see that requirement in the document

    <burn> milan: does not say "separately from the grammar"

    <burn> bjorn: you're right

    Milan: Are we using notes as reference or the formal Requirements as
    the basis

    Burn: You are right..the requirements document to be the basis, and
    not the notes

    Bjorn: Michael suggested that we use the text from the notes to
    reword the requirements doc

    <burn> milan and bjorn: we need text for the new requirements

    2. App. should be able to specify lang for each recognition

    Milan: My concern is that we are NOT capturing all our discussions
    in the Req. Spec

    Bjorn: we should not do this on the phone, but, let Michael
    incorporate the notes into the spec

    <burn> action for michael: add text for each new requirement

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

    Burn: to send Michael action item to update the spec incorporating
    the notes from minutes

    <marc> ACTION: Michael to add text for each new requirement
    [recorded in

    <trackbot> Sorry, amibiguous username (more than one match) -

    <trackbot> Try using a different identifier, such as family name or
    username (eg. mbodell2, mjohnsto)

    Michael is Michael Bodell

    <marc> ACTION: mbodell2 to add text for each new requirement
    [recorded in

    <trackbot> Created ACTION-1 - Add text for each new requirement [on
    Michael Bodell - due 2010-12-09].

    Burn: So, we do have agreement on the 2 new reqs for R2


      [24] http://lists.w3.org/Archives/Public/public-xg-htmlspeech/2010Dec/0017.html

    Burn: Few more emails came on this topic post my sending the summary
    earlier today
    ... Implementations must support SSML

    Agreed to drop R10

: R20

      [25] http://lists.w3.org/Archives/Public/public-xg-htmlspeech/2010Dec/0021.html

    Bjorn: Selecting a voice can be done without SSMl

    being a requirement


      [26] http://lists.w3.org/Archives/Public/public-xg-htmlspeech/2010Dec/0016.html

    Marc: Is it possible to indicate preference of voices ( like take
    female voice from among the available)

    Bjorn: But you cannot specify a specific engine from a specific
    vendor, is that correct?

    Burn: We have a req. to use SSML but not SSML 1.1,
    ... SSML1.1 has many of the features for selection of language etc.

    Bjorn: If there is no addtional burden, then requiring SSML1.1 seems
    ... If there is no addtional burden, then requiring SSML1.1 seems OK

    Burn: So, we can put it into the req. and deal with at the time of

    Marc: May be we should delay decision on this..

    <marc> Proposed wording:

    Marc: letting author to specify the language..Marc to type the

    <marc> Web application authors should be able to specify the voice
    with which some text is to be spoken.

    burn: That wording is OK by me

    <bringert> Web apps should be able to specify which voice is used
    for TTS

    Burn: Bjorn's text will replace the text for R20
    ... Team agreed


      [27] http://lists.w3.org/Archives/Public/public-xg-htmlspeech/2010Dec/0019.html



      [28] http://lists.w3.org/Archives/Public/public-xg-htmlspeech/2010Dec/0016.html

    <burn> there seemed to be consensus to add "When speech input is
    used to provide input to a web app, it should be possible for the
    user to select alternative input methods."

    I agree

    scribe: there is group agreement on this.

    <burn> for the next piece: Chan's last email was

      [29] http://lists.w3.org/Archives/Public/public-xg-htmlspeech/2010Nov/0207.html

    Burn: Burn: There is a second thread, initiated by Chan...since he
    is not on the call today, we can defer the detailed discussion on
    ... Chan has multiple threads...on this..so, we can just let this
    happen through email rather than on the phone

    Group has no objection with Dan Burnett's suggestion


      [30] http://lists.w3.org/Archives/Public/public-xg-htmlspeech/2010Dec/0019.html

    UNKNOWN_SPEAKER: webapp developer should be able to specify
    name-specific Language Model

    Burn: We can discuss, but check with Robert, as he has some comments
    on this

    Dahl: The comment from Robert came just last night..and afraid many
    did not have a chance to review it adequately

    <ddahl> actually, the requirement was only sent to the list for
    discussion last night

    Burn: Let's not discuss MUST vs. SHOULD now,.but if there are any
    other suggestions..we can discuss that..
    ... There is no standard for that right now..

    Bjorn: If the standard becomes available in future, it may be
    possible to specify this with URI

    Burn: It seems that we have agreement on this among the people on
    the call.....
    ... We will keep R12 as stated and confirm it on the mailing list


      [31] http://lists.w3.org/Archives/Public/public-xg-htmlspeech/2010Dec/0020.html

    <Milan> Transport layer security (e.g. HTTPS) if requested by the
    web app.

    UNKNOWN_SPEAKER: Milan to post the 3 requirements under this topic

    Bjorn: We agreed on the priority level of req..

    <smaug_> bringert: did you really said that ^^

    Burn: There are many reqs.. that may be very relevant...

    Bjorn: WebApp API and UA-2-SpeechAPI are two separate reqs.
    ... Correction on Bjorns' comment on priority level...Expunge from
    ... Core needed for any speech API and the other for WebApps..

    Burn: Is there anything else that we need to cover today?
    ... suggest Bjorn correct the statements himself with the
    appropriate wording..

    <bringert> My position on grouping requirements: We should split the
    requirements document in two sections: 1. Requirements needed for
    any HTML Speech API, 2. Requirements only needed for web app
    specified network speech services.

    End Minutes

    <bringert> What I said about priorities: There was a discussion of
    the relative priorities of the core Web app - UA API, and the
    network speech services API, and we didn't come to a conclusion.

    <bringert> My position (as I've communicated on the mailing list) is
    that the Web app API should be considered the core API, and the APIs
    for specifying and communicating with network speech services should
    be considered extensions of lower priority.

    <burn> s/Next Topic:/Topic:/g

Summary of Action Items

    [NEW] ACTION: mbodell2 to add text for each new requirement
    [recorded in
    [NEW] ACTION: Michael to add text for each new requirement [recorded

    [End of minutes]
Received on Thursday, 2 December 2010 22:21:36 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 20:16:48 UTC