[minutes] 10 November 2011

Group,

The minutes from today's call are available at http://www.w3.org/2011/11/10-htmlspeech-minutes.html

For convenience, a text version is embedded below.

Thanks to Patrick Ehlen for taking the minutes.

-- dan

**********************************************************************************
              HTML Speech Incubator Group Teleconference

10 Nov 2011

   [2]Agenda

      [2] http://lists.w3.org/Archives/Public/public-xg-htmlspeech/2011Nov/0058.html

   See also: [3]IRC log

      [3] http://www.w3.org/2011/11/10-htmlspeech-irc

Attendees

   Present
          Dan_Burnett, Dan_Druta, Glen_Shires, Olli_Pettay,
          Michael_Bodell, Debbie_Dahl, Patrick_Ehlen, Charles_Hemphill,
          Satish_Sampath, Michael_Johnston, Avery_Bishop

   Regrets
   Chair
          Dan_Burnett

   Scribe
          Patrick_Ehlen

Contents

     * [4]Topics
         1. [5]Review of updated final report draft
         2. [6]Issue 1
         3. [7]Moving explanation of "final" to end of section
         4. [8]Protocol issue 3
         5. [9]API issue 3: Updated re-reco example to use Robert's new
            version
         6. [10]API issue 4: Wrote up the EMMA mappings. This involved
            a new section explaining the mappings and an example to
            illustrate it
         7. [11]API Issue 5: Brought back the grammar/parameter
            functions. This involved uncommenting the
            functions/explanations, adding a new one for grammarString,
            and renaming and rewording appropriately.
     * [12]Summary of Action Items
     _________________________________________________________


Review of updated final report draft

   <burn> Report draft is in this email:
   [14]http://lists.w3.org/Archives/Public/public-xg-htmlspeech/2011Nov
   /0061.html

     [14] http://lists.w3.org/Archives/Public/public-xg-htmlspeech/2011Nov/0061.html

   Dan: Manually merged changes instead of applying add'l edits

   Issue 1

Issue 1

   Michael: Change in IDL in sec 7.1.4. resultDeleted event
   ... There's also an event for this (SpeechInputResultDeleted).

   <mbodell> link to the section with most of the changes:
   [15]http://www.w3.org/2005/Incubator/htmlspeech/finalreport/XGR-html
   speech.html#speechinputrequest-section

     [15] http://www.w3.org/2005/Incubator/htmlspeech/finalreport/XGR-htmlspeech.html#speechinputrequest-section

Moving explanation of "final" to end of section

Protocol issue 3

   Added Result-Index header field

   burn: changes are in a few different places; 1st in reco headers
   section.
   ... now adds a # of 1-20 digits, and added a description

API issue 3: Updated re-reco example to use Robert's new version

   all agree example looks good

   <mbodell> Link to new section on EMMA mappings:
   [16]http://www.w3.org/2005/Incubator/htmlspeech/finalreport/XGR-html
   speech.html#emma-to-interpretation

     [16] http://www.w3.org/2005/Incubator/htmlspeech/finalreport/XGR-htmlspeech.html#emma-to-interpretation

API issue 4: Wrote up the EMMA mappings. This involved a new section
explaining the mappings and an example to illustrate it

   michael: explains mapping section and example added
   ... in section 7.1.5

   <burn> this should be a static dated version of the current doc:
   [17]http://www.w3.org/2005/Incubator/htmlspeech/finalreport/XGR-html
   speech-20111110.html

     [17] http://www.w3.org/2005/Incubator/htmlspeech/finalreport/XGR-htmlspeech-20111110.html

   smaug: Specify whether all the emma namespace stuff in emma spec is
   required

   <mbodell> link to emma element: [18]http://www.w3.org/TR/emma/#s3.1

     [18] http://www.w3.org/TR/emma/#s3.1

   can probably do just emma specification but leave out location and
   xsi

   ddahl1: does not appear to be required

   charles: looks like xmlns:emma only can be used

   ddahl1: and the emma:emma version

   michael will drop the 3 lines assoc. with xsi stuff

API Issue 5: Brought back the grammar/parameter functions. This
involved uncommenting the functions/explanations, adding a new one for
grammarString, and renaming and rewording appropriately.

   <mbodell> Back to
   [19]http://www.w3.org/2005/Incubator/htmlspeech/finalreport/XGR-html
   speech.html#speechinputrequest-section

     [19] http://www.w3.org/2005/Incubator/htmlspeech/finalreport/XGR-htmlspeech.html#speechinputrequest-section

   Brought back some of the API grammar functions

   satish: seems fine

   burn: On issue 5, there was another resolution on built-in default;
   still needs to be done?

   mbodell: yes

   satish: needs array manipulation funcs?

   burn: better to leave the way it is now; can revisit delete
   capability later on

   smaug: misleading names on 7.1.4.2

   e.g., "Speech Input Request" (w/ spaces)

   mbodell will do that

   Topic; API Issue 6: Shortend the naming of resultEMMAXML and
   resultEMMAText to no longer have the result. They would nearly
   always be referenced from result property anyways so this just
   changes result.resultEMMAText to result.EMMAText)

   no objections

   <glen> Add clarification at end of description of AddGrammarElement
   method. "Note that grammars can be removed through direct
   maniplulation of the grammar array."

   API Issue 10: Added text property to TTS with explanation and also
   included statement about xml:lang

   <mbodell> See section:
   [20]http://www.w3.org/2005/Incubator/htmlspeech/finalreport/XGR-html
   speech.html#tts-section

     [20] http://www.w3.org/2005/Incubator/htmlspeech/finalreport/XGR-htmlspeech.html#tts-section

   no requests for changes around text attribute

   burn: Section 7.1.3 language
   ... language on "should use base64 encoding" is ambiguous
   ... people can probably figure out when they need to use base64 in
   UA
   ... those are all of the changes that have been applied so far
   ... will have more time in next few days to work on it
   ... practical issue: last date for call is next week
   ... we could (a) end call now and make more changes; or (b) try to
   do them live now during the call
   ... could also describe how everything gets done and reviewed

   MichaelJ: discussed EMMA mapping yet?

   <mbodell> e,,a at
   [21]http://www.w3.org/2005/Incubator/htmlspeech/finalreport/XGR-html
   speech.html#emma-to-interpretation

     [21] http://www.w3.org/2005/Incubator/htmlspeech/finalreport/XGR-htmlspeech.html#emma-to-interpretation

   MichaelJ: wasn't sure about "Boston flight 311" and then
   interpretation being "Boston"; could probably come up with more
   clear string->meaning mapping
   ... What is assumption on what ends up in interpretation?
   ... So a DOM object is hanging off interpretation?

   mbodell: is "any"; could be EMMA literal, or XML DOM document
   ... it's all javascript

   MichaelJ: Other case is where it's a JSON payload; so interpretation
   would contain JSON?

   mbodell: didn't have any mapping for that; only literals and xml

   MichaelJ: so we may want to consider having JSON within literal in
   CDATA

   mbodell: not CDATA; just a string

   <Michael> <emma:literal><![CDATA[
   {_cmd:"CALL",_obj:"joe_bloggs"}]]></emma:literal>

   mbodell: these will be normal javascript properties
   ... based on what's there, if you have emma:literal, you just get a
   string with that value

   MichaelJ: but there are attribs in EMMA 1.0 that describe what type
   of data you have

   mbodell: if you want to eval your content, you can do it in
   javascript

   MichaelJ: that's reasonable. could modify this down the line, but
   you can always eval or parse it

   <mbodell> For context the description of sting literal in emma is:
   [22]http://www.w3.org/TR/emma/#s3.5

     [22] http://www.w3.org/TR/emma/#s3.5

   burn: open again for use of remaining time

   ddahl1: is there a deadline for sending comments to list, and how
   should time be allocated in general?

   burn: editing references can wait until end (or even after)
   ... if anyone had updates for examples, send them to list
   ... a few other issues that require more substantive text:
   ... such as text around issue 11 on bindings
   ... understand that more work needs to happen here and there are
   gotchas

   11, 17, 18 related

   mbodell will take a crack at that

   burn: michael was also going to take issue 10

   Michael will also do 23

   22 still needs to be done

   burn will do 22

   mbodell already did 10

   burn: 8 and 13 are on switching to collections from arrays
   ... good example from Satish on making it look like FileList

   MichaelJ: sent example w/ literal w/ semantics

   burn: aligning adjustment of requirements

   burn will take that one

   burn: how do we wrap up effectively?

   <Avery> Sorry I'm late

   burn: can contentful changes be done by, say, tuesday?
   ... gives time for review by thursday
   ... other option is a very slight extension to beginning of december

   mbodell: let's just aim to get content in by tuesday
   ... other changes can go over email

   burn: would prefer that as well
   ... sounds like it's just a matter of getting those changes in
   ... next and final call is in one week

   <Avery> I might be able to make it Thursday

Received on Thursday, 10 November 2011 20:38:28 UTC