RE: feeback on the reco element

Hello Kenny,

Just wanted to give you a heads up that the HTML Speech XG project is in a transitional stage.  We're trying to find a new home for the work as our charter has expired.  As such I don't know if anyone will be actively looking at the suggestions you have raised.  They might be queued for discussion once the group has landed.

I will put out the quick comment that I don't think the declarative (tag-based) API received as much focus during the XG process as the procedural model.  I could be wrong, but I think the folks at Google are pushing to get rid of this altogether and if so I would personally support them in this effort.

Hope to be working with you soon.



-----Original Message-----
From: Kang-Hao (Kenny) Lu [mailto:kennyluck@csail.mit.edu] 
Sent: Thursday, January 05, 2012 5:58 AM
To: HTML Speech XG
Subject: feeback on the reco element

1. There should be a proper example about the @for attribute for <reco>

The first example in 7.1.7:

 <reco for="q"/>
 <input id="q" name="q" type="text"/>

could be better written as

 <reco>
   <input name="q" type="text"/>
 </reco>

It would be better to have an example that shows why @for is useful (and why nested <reco> hasn't covered the use cases of @for already)

2. The IDL lang attribute on HTMLRecoElement inherited from SpeechReco has conflict with the lang attribute reflecting the @lang attribute.

Is unifying the two the idea? If it is, I am not sure having a separate request attribute is making sense. Is it the intention to have reco.lang === reco.request.lang and so on?

3. I am not sure if NamedConstructor=Reco(in DOMString for) is good design.

Why isn't NamedConstructor=Reco(in HTMLElement form) better? FormData[1] from XMLHttpRequest uses something like this too.

4. In the description of the control attribute, the third paragraph says "The reco IDL attribute", which should probably read "The recos IDL attribute". Also, it would be less confusing if the second and third paragraphs are merged.

[1] http://dvcs.w3.org/hg/xhr/raw-file/tip/Overview.html#interface-formdata



I think in general, the sections of the document related to <reco> and <tts> should have more cross-document references to the HTML spec for terms like "active document" and so on. The intention to make <reco> a form-associated element should be written down somewhere in this section too.


Cheers,
Kenny

Received on Friday, 6 January 2012 19:36:37 UTC