Re: On test suites

I see multiple tracks here:

   1. Speech recognition part of the spec has been progressing well and the
   remaining pieces seem to be
      1. the attributes in discussion (confidenceThreshold, maxNBest, ..)
      2. the use cases with example EMMA xml.
   2. Speech synthesis needs more discussion. To start CG members should
   review the proposed API and suggest changes as necessary. There were
   earlier discussions about
      1. adding more events to TTS
      2. SSML support
      3. features like enqueueing next utterance while current one is going
      on

Starting separate discussions around these with proposed API changes would
be great.

Cheers
Satish


On Tue, Jun 19, 2012 at 4:13 PM, Young, Milan <Milan.Young@nuance.com>wrote:

> I'm in agreement with what Satish has written below (and the follow-ups
> from Jim and Ian).  I'd like confirmation from the chair that this is the
> official position of the group.
>
> I'd also like to see a roadmap that gets us to TPAC with a stable first
> edition in hand.  I will be happy to take a stab at that if nobody else
> does.
>
> Thanks
>
>
> -----Original Message-----
> From: Ian Jacobs [mailto:ij@w3.org]
> Sent: Tuesday, June 19, 2012 6:32 AM
> To: Jim Barnett
> Cc: Satish S; Young, Milan; Glen Shires; jerry@jerrycarter.org;
> schepers@w3.org; olli@pettay.fi; bringert@google.com; raj@openstream.com;
> dahl@conversational-technologies.com; public-speech-api@w3.org
> Subject: On test suites
>
>
> On 19 Jun 2012, at 7:57 AM, Jim Barnett wrote:
>
> > Satish,
> > I don't think that there are _any_ requirements for going to a WG, other
> than people being willing to sign up for it.  You're right that test suites
> are required only rather late in the WG process.
>
> Hi Jim,
>
> Just a note that while test suites are definitely encouraged, they are not
> required by the W3C process. They do provide a good way of determining two
> interoperable implementations all all the features, which is the
> requirement.
>
> Ian
>
>
> >
> > -          Jim
> >
> > From: Satish S [mailto:satish@google.com]
> > Sent: Tuesday, June 19, 2012 7:24 AM
> > To: Young, Milan
> > Cc: Glen Shires; Jim Barnett; jerry@jerrycarter.org; ij@w3.org;
> > schepers@w3.org; olli@pettay.fi; bringert@google.com;
> > raj@openstream.com; dahl@conversational-technologies.com;
> > public-speech-api@w3.org
> > Subject: Re: Co-chair
> >
> > Once the speech API spec gets adopted by a WG I think ongoing work can
> happen there and this CG's purpose would be fulfilled. So I would think
> subsequent revisions happen in the WG.
> >
> > Re: TPAC, that gives us about 4 months from now and looks like
> sufficient time to cover major topics including TTS.
> >
> > Re: test suites, that is a requirement when the spec goes to a Candidate
> Recommendation and I believe we can take the spec to a WG before the test
> suite is ready. Glen, please correct me if I'm wrong.
> >
> > Cheers
> > Satish
> >
> >
> > On Tue, Jun 19, 2012 at 2:23 AM, Young, Milan <Milan.Young@nuance.com>
> wrote:
> > Inline...
> >
> >
> > From: Glen Shires [mailto:gshires@google.com]
> > Sent: Thursday, June 14, 2012 8:13 AM
> > To: Young, Milan
> > Cc: Jim Barnett; jerry@jerrycarter.org; ij@w3.org; schepers@w3.org;
> > olli@pettay.fi; bringert@google.com; satish@google.com;
> > raj@openstream.com; dahl@conversational-technologies.com;
> > public-speech-api@w3.org
> > Subject: Re: Co-chair
> >
> > Yes, I believe we need to complete the initial version of this spec as a
> CG before formally proposing to a WG that it be put on a standards track.
>  By initial version, I mean a version that supports the majority of use
> cases, but it also implies that there will be subsequent versions that add
> additional features.  By keeping the initial version simple, and avoiding
> bloat, we make it easier for WG to take on the work, and for multiple
> browser vendors to implement.
> >
> > [Milan] Could you please clarify whether work on those subsequent
> revisions would happen in this CG vs a WG?
> >
> >
> > This is consistent with "The goal and scope of this Community
> > Group...For this initial specification, we believe that a simplified
> > subset API will accelerate implementation, interoperability testing,
> > standardization and ultimately developer adoption." [1] [2]
> >
> > Based on our substantial progress so far and taking into account the
> pending work (TTS, test suites, pending topics, finer aspects of the API
> that may come up) I estimate the initial version of this spec will be
> completed by end of the year. If we can resolve key topics quickly, the
> spec can be ready sooner.
> >
> > [Milan] My preference is that we have a draft worthy of review by the
> next TPAC.  I believe this will help inform our decision on joining an
> existing WG or creating our own.
> >
> > In order to do that, we'll need a better gauge on the topics ahead.
>  Would you be available to sketch out our timeline?  Of particular concern
> to me is your mention of "test suites".  That seems like a topic that could
> drag on for quite some time.  Do you view this as a prerequisite for
> inclusion into a WG?
> >
> > / Milan
> >
> >
> > Glen
> >
> > [1]
> > http://lists.w3.org/Archives/Public/public-new-work/2012Apr/0000.html
> > [2] http://www.w3.org/community/speech-api/
> >
> >
> > On Thu, Jun 14, 2012 at 12:39 AM, Young, Milan <Milan.Young@nuance.com>
> wrote:
> > While I agree that we need to move this effort to a WG, we must be
> careful not to splinter.
> >
> > The truth is that this community has made progress cleaning up the scope
> of the XG report.  We also should keep in mind that a principle reason
> we're in this CG instead of some of the more attractive WGs like WebApps is
> because we lack consensus.  Starting a new WG while the CG is still in
> progress will not impress anyone.
> >
> > Glen, I would like to know your vision and timeline for the transition.
> >
> >
> > From: Jim Barnett [mailto:Jim.Barnett@genesyslab.com]
> > Sent: Thursday, June 14, 2012 12:11 AM
> > To: Young, Milan; gshires@google.com; jerry@jerrycarter.org;
> > ij@w3.org; schepers@w3.org
> > Cc: olli@pettay.fi; bringert@google.com; satish@google.com;
> > raj@openstream.com; dahl@conversational-technologies.com;
> > public-speech-api@w3.org
> > Subject: Re: Co-chair
> >
> > I think that we can move to a standards-track group at any time. The
> main thing that we need to do is to submit a charter, first to W3C
> management and then to the AC list. Dan has a draft charter, I think, that
> can serve as a template. Once we agree on the content, we submit it, handle
> any comments we get, and we're in business.
> >
> > Jim
> >
> > From: Young, Milan <Milan.Young@nuance.com>
> > To: Glen Shires <gshires@google.com>; Jerry Carter
> > <jerry@jerrycarter.org>; Ian Jacobs <ij@w3.org> (ij@w3.org)
> > <ij@w3.org>; Doug Schepers (schepers@w3.org) <schepers@w3.org>
> > Cc: olli@pettay.fi <olli@pettay.fi>; Jim Barnett; bringert@google.com
> > <bringert@google.com>; satish@google.com <satish@google.com>;
> > raj@openstream.com <raj@openstream.com>;
> > dahl@conversational-technologies.com<dahl@conversational-technologies.
> > com>; public-speech-api@w3.org <public-speech-api@w3.org>
> > Sent: Wed Jun 13 16:38:19 2012
> > Subject: RE: Co-chair
> >
> > Taking a step back, we're in a situation where a Google representative
> decides when consensus is reached, and if we lack consensus we default to
> whatever Google wanted earlier.  Do the folks in this community feel this
> is a path to building a spec that has the broad-based support needed to
> attract missing browser and speech vendors?
> >
> > I'd also like to call out an recent instance where consensus was
> reached, but the agreed changes did not make their way into the spec.  This
> happened near the end of the EMMA thread where Satish, Deborah, and I
> finally agreed to drop the requirement for EMMA attributes in exchange for
> adding use cases [1].  But when the changes were pushed through, they were
> missing the compromise text [2].  And my notification to this problem
> didn't generate any response from the chair or editors [3].  This is
> especially worrisome given that we just published our first draft (sans
> compromise text) without any advanced notification, vote, or opportunity
> for review [4].  Perhaps this is simply a case of broken timeline
> expectations, but given that my requests have fallen off the proverbial
> radar several times before (most recently [5]), it feels like a bias is at
> play.
> >
> > I would like to hear from others in the community on this topic.  I'm
> particularly interested to know thoughts around the formation of an
> official WG where we can produce a standards-track specification.
> >
> > Thanks
> >
> > [1]
> > http://lists.w3.org/Archives/Public/public-speech-api/2012Jun/0060.htm
> > l [2]
> > http://lists.w3.org/Archives/Public/public-speech-api/2012Jun/0061.htm
> > l [3]
> > http://lists.w3.org/Archives/Public/public-speech-api/2012Jun/0062.htm
> > l [4]
> > http://lists.w3.org/Archives/Public/public-speech-api/2012Jun/0076.htm
> > l [5]
> > http://lists.w3.org/Archives/Public/public-speech-api/2012Jun/0010.htm
> > l
> >
> >
> > From: Glen Shires [mailto:gshires@google.com]
> > Sent: Wednesday, June 13, 2012 8:02 AM
> > To: Jerry Carter
> > Cc: Young, Milan; olli@pettay.fi; Jim Barnett; bringert@google.com;
> > satish@google.com; raj@openstream.com;
> > dahl@conversational-technologies.com; public-speech-api@w3.org
> > Subject: Re: Co-chair
> >
> > Changes to the spec and to the structure of this CG are decided by rough
> consensus. There is no clear consensus on the co-chair proposal, so there
> will be no changes in the structure of this CG at this time.
> >
> > Glen Shires
> >
> >
> >
> >
>
> --
> Ian Jacobs (ij@w3.org)    http://www.w3.org/People/Jacobs/
> Tel:                                      +1 718 260 9447
>
>