15:59:34 <RRSAgent> RRSAgent has joined #pf
16:01:48 <chaalsNCE> chaalsNCE has changed the topic to: PF working group call: http://www.w3.org/WAI/PF/#meetings
16:03:45 <Zakim> +[Gallaudet]
16:03:51 <Zakim> +Scott
16:05:26 <chaalsNCE> agenda+ VoiceXML
16:05:35 <chaalsNCE> agenda+ XAG triage
16:05:50 <chaalsNCE> agenda+ 2 minutes
16:06:31 <chaalsNCE> Present - Scott McGlashan - Voice
16:06:38 <chaalsNCE> - Al Gilman - PFWG
16:06:50 <chaalsNCE> - Kerry Barbin
16:07:23 <chaalsNCE> (invited expert - technologist at Gallaudet University)
16:07:40 <chaalsNCE> - Charles McCathieNevile - PFWG
16:09:19 <Al> Al has joined #pf
16:09:36 <Zakim> +??P10
16:10:44 <chaalsNCE> - Lisa Seeman
16:12:05 <chaalsNCE> zakim, agenda?
16:12:06 <Zakim> I see 3 items remaining on the agenda:
16:12:07 <Zakim> 1. VoiceXML [from chaalsNCE]
16:12:08 <Zakim> 2. XAG triage [from chaalsNCE]
16:12:09 <Zakim> 3. 2 minutes [from chaalsNCE]
16:16:24 <Zakim> +??P12
16:16:57 <chaalsNCE> - GotteFried Zimmermann - PFWG
16:17:15 <chaalsNCE> (WHoops. Gottfried not Gottefried)
16:18:44 <Gottfried> Gottfried has joined #pf
16:19:09 <chaalsNCE> [Al explains the use of IRC for logging...]
16:19:36 <chaalsNCE> zakim, take up agenda 3
16:19:37 <Zakim> agendum 3. "2 minutes" taken up [from chaalsNCE]
16:20:14 <chaalsNCE> CMN: Discovered that my response to the document Jim et al produced has disappeared into cyberspace without a trace <sigh/>
16:20:35 <Al> CMN wrote a response to draft doc.
16:20:46 <Al> ACTION CMN rewrite response
16:20:48 <chaalsNCE> ACTION: CMN rewrite my response to the document on Voice / Accessibility
16:21:27 <chaalsNCE> CMN wrote a VoiceXMl issues page
16:22:08 <Al> There is a thread on WAI-XTECH touching on XLink in the XAG
16:22:38 <chaalsNCE> [http://www.w3.org/WAI/PF/Group/voice.html | PFWG's voice issues page ]
16:22:54 <Zakim> + +1.503.645.aabb
16:22:58 <chaalsNCE> Hotbuttons for today; XAG triage
16:23:18 <chaalsNCE> - Jim Larson
16:24:34 <chaalsNCE> (voice)
16:25:09 <chaalsNCE> Scott: Working with VoiceXML, on Version 2.0
16:25:23 <chaalsNCE> Scott: dealing with Last call issues on VoiceXML 2
16:25:51 <chaalsNCE> ... hot buttons for today: what do we need to change in current spec?
16:25:58 <chaalsNCE> ... (before next week...)
16:26:14 <chaalsNCE> Lisa: no hot buttons today.
16:27:06 <chaalsNCE> ... writing up issues of synthesising hebrew / arabic (because vowels are not necessarily written down, so words can be ambiguous)
16:27:21 <chaalsNCE> ... trying to get a tool developer for clear language work
16:28:09 <chaalsNCE> GZ: Working on NCITS V2 - a US project for using a personal remote control for devices
16:28:34 <chaalsNCE> ... hot buttons: What are the changes we could make and what are the ones we need to hold off for the future?
16:28:51 <chaalsNCE> ... (dream: How to get to single authoring techniques)
16:30:14 <chaalsNCE> ... There is a workshop onn this topic in Germany end of September http://www.w3.org/2002/07/DIAT/
16:31:46 <chaalsNCE> zakim, take up agendum 1
16:31:47 <Zakim> agendum 1. "VoiceXML" taken up [from chaalsNCE]
16:32:01 <chaalsNCE> Jim: think we should be talking about requirements on Voice Browsers
16:32:13 <chaalsNCE> Scott: THink we should concentrate on VoiceXML 2.0 spec
16:32:24 <chaalsNCE> Jim: saw at least 3 requirements for browsers:
16:32:42 <chaalsNCE> 1. protocol to have a text-based output from a voice server
16:32:53 <chaalsNCE> 2. protocol to have a text-based input
16:33:06 <chaalsNCE> 3. Universal commands - change volume, change speech rate, ...
16:33:48 <chaalsNCE> Scott: Relating it to VoiceXML: allowing text in/out can be acchieved by SGRS which is required.
16:34:15 <chaalsNCE> ... SSML output is SSML which could be sent as text instead of speech, with the exception of the audio element.
16:34:49 <chaalsNCE> ... WAI comments says that text equivalents should be required as part of the format. (If there is no mechanism, there will be no alternative text)
16:35:11 <chaalsNCE> ... suggestion is to add an alt attribute for audio, like the HTML alt
16:35:15 <chaalsNCE> Jim: Agree
16:35:21 <chaalsNCE> ... put taht in guidelines
16:35:30 <chaalsNCE> Scott; Right, but there still needs to be the element in the language
16:35:35 <chaalsNCE> q+
16:36:16 <chaalsNCE> Jim: add a fourth bullet to number one that says include an alt attribute
16:36:43 <chaalsNCE> Al: would you be willing to make that "include a mechanism"? Our current thinking is that an element is more flexible and useful.
16:37:35 <chaalsNCE> Jim: Can say text equivalent instead of "alt attribute"
16:37:48 <chaalsNCE> Scott: We have to produce a spec, so we should address the issue now if posible
16:39:15 <chaalsNCE> CMN: Making the audio element have content is one approach.
16:39:30 <chaalsNCE> ... there may be times when the audio isn't necessary and people want it to go away.
16:39:49 <chaalsNCE> Scott: There is text content as a fallback when the audio is not avilable
16:39:57 <Al> q+ al
16:40:23 <chaalsNCE> Scott: We looked at ruby and found it inadequate for voice technology - why is it beter than plain text
16:41:06 <chaalsNCE> CMN: "when the audio is not available" seems to match exaactly "when I can't hear the audio"
16:42:20 <chaalsNCE> ... ruby seems to match the say-as construct in SSML - it is more flexible, which has up and down sides.
16:42:56 <chaalsNCE> Lisa: When we have a text equivalent we may need guidelines about writing those - using unambiguous language, not acronyms, ...
16:43:12 <chaalsNCE> ... Is it possible to require the alternative in the spec?
16:43:57 <chaalsNCE> Al: The point isn't that Ruby is better than SSML but that elelment is better than attribute because you can have diffedrent kinds of markup.
16:44:48 <chaalsNCE> ... our first things would be - "prove that the fallback is a different requirement than the accessibility requirement, since the same information seems requireed in both" and we want to implement universal design and avoid special-cases where possible.
16:45:05 <chaalsNCE> Kerry: what is Ruby?
16:45:39 <chaalsNCE> Al: Ruby is an XMl construct where you can place phoetic script (for example) and the ideographic script. It is a common typographical convention in East Asian languages
16:46:14 <chaalsNCE> e.g. <ruby><rb>W3C</rb><rt>World Wide Web Consortium</rt></ruby>
16:46:37 <chaalsNCE> Scott: We can take this back and see if there is a difference. I can't think of the arguments but it seemed previously that they were differnt.
16:46:46 <chaalsNCE> ... for discussion on tuesday
16:47:07 <chaalsNCE> Jim: If we believe that an element is better than an attribute, what do we call it?
16:47:16 <chaalsNCE> Al: The content of Audio.
16:47:36 <chaalsNCE> Scott: Issue with that: The content can itself be another audio.
16:48:25 <chaalsNCE> Al: There is a thread dealin with the object element in HTML - the legacy is you use the outermost element that can be applied. Looking to be able to pick the most appropriate one.
16:49:24 <chaalsNCE> CMN: Don't see a problem with multiple audio elements being nested. The problem is that we cannot test for presence of "useful" content, but that's a tradeoff for the flexibility of content
16:50:18 <chaalsNCE> Scott: Agreed that we need text equivalents for audio elements. PFWG has a strong preference for it being element content. Question whether it should be seperated from other content of audio - OFWG has a preference for it being the same.
16:50:43 <chaalsNCE> ... which would use the same content for accessibility and fallback, until shown that there are different fuctions for those cases.
16:51:04 <chaalsNCE> ... At the moment what should that be? Voice is suggesting SSML, PFWG is suggesting Ruby.
16:51:37 <chaalsNCE> Al: ruby is incidental - we are not tied to it in particular, just the ability to include XML markup not just attribute content.
16:51:55 <chaalsNCE> ... intersted in working on what can go in there and work.
16:52:26 <chaalsNCE> Having a flexible namespace to allow linking to another namespace is cool - allows linking to things like symbolic languages.
16:52:36 <chaalsNCE> (that was Lisa)
16:52:57 <chaalsNCE> Lisa: Ruby allows multiple renderings which is useful beyond just japanese.
16:53:09 <chaalsNCE> ... and ruby is already built...
16:55:33 <chaalsNCE> CMN not worried about having Ruby. If we have XML content we can use SSML for now and later used mixed namespaces to produce desirted formats by XSLT if there are differing implementations
16:56:05 <chaalsNCE> Scott: Appendix H on accessibilty. At the time we wrote that we had understtod that people would make different version of content for different media. Would like someone to review the appendix.
16:56:33 <chaalsNCE> Al: PFWG should take an action item to propose alnguage.
16:57:10 <chaalsNCE> Scott; that would be great. Deadline of 18 Septmeber?
16:57:35 <chaalsNCE> ACTION: PFWG to make a proposal for accessibiltiy appendix. Deadline 18 September.
16:57:54 <chaalsNCE> Al. There are some things where I don't think we change the spec, but want a record of the reasoning.
16:58:17 <chaalsNCE> ... 1. Is having content for the audio element something that should be a conformance requirement for documents?
16:58:36 <chaalsNCE> ... 2. Having functional equivalence of DTMF commands as a conformance requirement.
16:59:16 <chaalsNCE> ... these have to be addressed even if they don't result in spec changes.
16:59:33 <chaalsNCE> Jim: number 2 is listed as a task for authoring guidelines
16:59:48 <chaalsNCE> Al: There is a possibility for saying that it should be a formal validity requirement.
17:00:26 <chaalsNCE> Jim: I suggest we write a document containing the arguments and motivations, based on the working paper sent previously http://lists.w3.org/Archives/Member/w3c-wai-pf/2002JulSep/att-0086/01-PFWG_VBWGv11.html
17:01:15 <chaalsNCE> CMN: including some user agent conformance requirements (speech rate, delay, volume) as a conformance requirement for user agents is a possibility
17:01:30 <chaalsNCE> Jim: Cannot at this point agree to making things mandatory on behalf of the group.
17:01:53 <Zakim> -Lisa
17:02:48 <chaalsNCE> Al: Whether things go into guidelines or specs is something we have to justify to each of our working groups
17:03:47 <chaalsNCE> Scott: There are two conformance sections - one about documents, one about user agents. User agents must already conform to the syntax and semantics of VoiceXML, SGRS and SSML
17:04:16 <chaalsNCE> ... there are outstanding change requests to provide control over volume etc within the language. These won't be in VoiceXML 2.0 because they can be done in user agents.
17:04:33 <chaalsNCE> Scott:
17:05:09 <chaalsNCE> Did you mean things are "should" requirements or "must" requirements?
17:05:17 <chaalsNCE> Al: I wanted "must"
17:05:24 <chaalsNCE> Scott: I had suggested "should"
17:06:14 <chaalsNCE> Scott; Understand the requirement on DTMF equivalence and text equivalents. there is some politics to be worked through the group...
17:06:53 <chaalsNCE> Scott: Think I have a clear picture of what we are looking at for text equivalents.
17:07:34 <chaalsNCE> Al: That is our number one. We have on our shopping list the browser requirements and DTMF... will need to have a story of why things aren't there.
17:07:48 <chaalsNCE> Scott: We are looking at making things for version 3.0 too....
17:08:05 <Al> I meant "must" on document, not just player
17:08:10 <chaalsNCE> Scott will discuss these issues on Tuesday at voice call and get back to us.
17:08:13 <Zakim> -Scott
17:08:59 <chaalsNCE> Jim: The big picture. I think it would be nice to have a single document that follows the outline of the above-mentioned working document.
17:09:22 <chaalsNCE> ... should lay out the issues and divide things in terms of whetehr they are guidelines, specification requirements, or research issues.
17:09:32 <chaalsNCE> ... is such a document a good idea?
17:09:52 <Al> q+ al
17:12:11 <chaalsNCE> CMN I think it is a good idea, and that a sumary of it would be good material for an appendix on accessibility. It will take at least some person-days, perhaps person-weeks.
17:12:51 <chaalsNCE> Al: I see different long-term homes for some of the topics. For example there is stuff that should end up being looked after by the Web Content Accessibiltiy Guidelines group's work.
17:13:21 <chaalsNCE> ... but they don't have the information. Can we do this fast enough to produce an integrated document, but recognise that bits of work wil live on and be developed in various areas?
17:13:33 <chaalsNCE> ... for example see some of this impacting multimodal interaction work.
17:14:11 <chaalsNCE> Jim: When we write this document we should indicate within the text who we think bits are relevant to besides us...
17:15:22 <chaalsNCE> Al: We need to have assigned resources from each side, and we need to talk to the other groups who we think we are giving this to.
17:15:41 <chaalsNCE> ... if this is not on a REC track but a fast-development document I think it is possible.
17:15:46 <chaalsNCE> Jim: Think that is quite reasonable.
17:15:57 <chaalsNCE> Al: need tto get my volunteers and think it should be a go.
17:16:26 <chaalsNCE> Jim: Should have 4 sections - intro, spec requirements, guidelines, future research. Can we agree on that?
17:18:30 <chaalsNCE> CMN I don't think we should divide the problems into categories like that to start - start out by gathering the issues and look at potential solutions - since there are different possible approaches for the same problem and it would help to avoid the argument over which is the right one.
17:19:07 <chaalsNCE> Jim: Thnk I agree - need to start by exploring the space. The last three sections are to sort the possible solutions and work them into real concrete ones.
17:19:09 <Gottfried> q+
17:19:57 <chaalsNCE> Al: Appears to me that if we draw the tree as two items - analysis of the problem and potential solutions (which can be divided into the three categories) it might work.
17:20:17 <chaalsNCE> ... as you develop the lower sections, we wil want to have cross-references to where one potentiall solution impacts on others.
17:21:15 <chaalsNCE> jim: Think the process of writing the document will clarify a lot of things.
17:21:56 <chaalsNCE> GZ: Wanted to back up Charles. Go along the problems. Some problems may cause actions on several catgories at once.
17:22:19 <chaalsNCE> Jim: That is certainly the case. Some problems are divided into sub-problems and some of tose have different approaches.
17:24:33 <chaalsNCE> zakim, close agendum 1
17:24:34 <Zakim> agendum 1 closed
17:24:35 <Zakim> I see 1 item remaining on the agenda:
17:24:36 <Zakim> 2. XAG triage [from chaalsNCE]
17:27:32 <chaalsNCE> CMN made proposal for dealing with issues: http://lists.w3.org/Archives/Member/w3c-wai-pf/2002JulSep/0141.html
17:28:31 <chaalsNCE> ... summary - defer Xlink issue (mark it in draft), mark "what modules are accessible" in draft as an issue, merge 1.3 and 2.9, add a technique to 1.3, rest as status quo / editorial.
17:28:50 <chaalsNCE> Al: Ok. Will push this on xtech, get resolutions in two weeks.
17:29:07 <chaalsNCE> ====
17:29:11 <chaalsNCE> Voice continued:
17:29:28 <chaalsNCE> Al: Any preferences on questions like shoudl chaals make a document on the website?
17:29:35 <chaalsNCE> Jim: sounds fine to me...
17:30:15 <Zakim> -Al
17:30:16 <Zakim> -Gottfried
17:30:18 <Zakim> -Jim
17:30:35 <Gottfried> Thank you, Charles, for your incredible job in taking notes, controling Zakim, and simultaneously contributing to the discussion.  I am just amazed!
17:30:48 <Zakim> - +33.4.92.38.aaaa
17:30:50 <Zakim> WAI_PF()12:00PM has ended
17:30:57 <chaalsNCE> ACTION: Chaals post new version of the problems and solutions document.
17:31:16 <chaalsNCE> Thanks Gottfried
17:31:31 <chaalsNCE> Actually anyone can drive Zakim...
17:31:36 <chaalsNCE> zakim, help
17:31:37 <Zakim> Please refer to http://www.w3.org/2001/12/zakim-irc-bot for more detailed help.
17:31:39 <Zakim> Some of the commands I know are:
17:31:40 <Zakim>  xxx is yyy       - establish yyy as the name of unknown party xxx
17:31:42 <Zakim>                     if yyy is 'me' or 'I', your nick is substituted
17:31:43 <Zakim>  xxx may be yyy   - establish yyy as possibly the name of unknown party xxx
17:31:45 <Zakim>  I am xxx         - establish your nick as the name of unknown party xxx
17:31:48 <Zakim>  xxx holds yyy [, zzz ...] - establish xxx as a group name and yyy, etc. as participants within that group
17:31:50 <Zakim>  xxx also holds yyy - add yyy to the list of participants in group xxx
17:31:51 <Zakim>  who's here?      - lists the participants on the phone
17:31:54 <Zakim>  who's muted?     - lists the participants who are muted
17:31:56 <Zakim>  mute xxx         - mutes party xxx (such that 60# will not work)
17:31:57 <Zakim>  unmute xxx       - reverses the effect of "mute" and of 61#
17:31:59 <Zakim>  is xxx here?     - reports whether a party named like xxx is present
17:32:01 <Zakim>  list conferences - reports the active conferences
17:32:02 <Zakim>  this is xxx      - associates this channel with conference xxx
17:32:03 <Zakim>  excuse us        - disconnects from the irc channel
17:32:04 <Zakim> I last learned something new on $Date: 2002/09/04 17:33:37 $
17:32:20 <chaalsNCE> zakim, bye
17:32:21 <Zakim> Zakim has left #pf
17:32:26 <chaalsNCE> rrsagent, bye
17:32:26 <RRSAgent> I see 3 open action items:
17:32:26 <RRSAgent> ACTION: CMN rewrite my response to the document on Voice / Accessibility [1]
17:32:26 <RRSAgent>   recorded in http://www.w3.org/2002/09/04-pf-irc#T16-20-48
17:32:26 <RRSAgent> ACTION: PFWG to make a proposal for accessibiltiy appendix. Deadline 18 September. [2]
17:32:26 <RRSAgent>   recorded in http://www.w3.org/2002/09/04-pf-irc#T16-57-35
17:32:26 <RRSAgent> ACTION: Chaals post new version of the problems and solutions document. [3]
17:32:26 <RRSAgent>   recorded in http://www.w3.org/2002/09/04-pf-irc#T17-30-57