- From: Olli Pettay <Olli.Pettay@helsinki.fi>
- Date: Thu, 11 Nov 2010 13:49:16 +0200
- To: Bjorn Bringert <bringert@google.com>
- CC: Michael Bodell <mbodell@microsoft.com>, Robert Brown <Robert.Brown@microsoft.com>, Dan Burnett <dburnett@voxeo.com>, "public-xg-htmlspeech@w3.org" <public-xg-htmlspeech@w3.org>
On 11/11/2010 01:17 PM, Bjorn Bringert wrote: > I think that specifying the language without a grammar is useful for > cases where the app isn't specifying a grammar at all (e.g. > dictation). Yes. > > A separate issue is what to do if the app requests a language that the > speech recognizer doesn't support. Maybe we should add another > requirement along the lines of: > > - Web apps should have access to a list of the languages supported by > the current speech service implementations. This might lead to fingerprinting, at least when using local speech engines, so we need to be very careful with it. It is a bit similar to getting keyboard locale, and that will be possible only during key events (per DOM 3 Events), not any random time to prevent fingerprinting. -Olli > > /Bjorn > > On Thu, Nov 11, 2010 at 8:00 AM, Michael Bodell<mbodell@microsoft.com> wrote: >> I mostly agree. The only wrinkle is if you are allowed multiple grammars and different languages are specified in each grammar what should happen? Should it fail? Should it do one recognition for each language and return a bunch of results? Do multiple recognition but return one result? Do one recognition in the language of one of the grammars and use the foreign language grammars in the "native" language picked? >> >> One can imagine specifying the language of the recognition separately from the grammar in question and then it either being an error if foreign grammars are specified, or else foreign grammars are ignored, or else you make a best effort with foreign grammars. >> >> This may get more into the proposal details and edge cases, but I think this is sort of what the R8 requirement gets at. >> >> -----Original Message----- >> From: public-xg-htmlspeech-request@w3.org [mailto:public-xg-htmlspeech-request@w3.org] On Behalf Of Robert Brown >> Sent: Wednesday, November 10, 2010 5:35 PM >> To: Dan Burnett; public-xg-htmlspeech@w3.org >> Subject: RE: R8. Web application must be able to specify language of recognition. >> >> This one feels redundant. SRGS specifies the language. And while an application will probably also need to specify a particular recognition resource based on the language it wants to recognize, this logic might be best left to the application. These requirements are already covered (SRGS, and specifying what resource to use). Are there other reasons for this requirement? >> >> -----Original Message----- >> From: public-xg-htmlspeech-request@w3.org [mailto:public-xg-htmlspeech-request@w3.org] On Behalf Of Dan Burnett >> Sent: Wednesday, November 10, 2010 3:24 PM >> To: public-xg-htmlspeech@w3.org >> Subject: R8. Web application must be able to specify language of recognition. >> >> Group, >> >> This is the next of the requirements to discuss and prioritize based on our ranking approach [1]. >> >> This email is the beginning of a thread for questions, discussion, and opinions regarding our first draft of Requirement 8 [2]. >> >> Please discuss via email as we agreed at the Lyon f2f meeting. >> Outstanding points of contention will be discussed live at the next teleconference. >> >> -- dan >> >> [1] http://lists.w3.org/Archives/Public/public-xg-htmlspeech/2010Oct/0024.html >> [2] http://lists.w3.org/Archives/Public/public-xg-htmlspeech/2010Oct/att-0001/speech.html#r8 >> >> >> >> >> >> > > >
Received on Thursday, 11 November 2010 11:49:59 UTC