Re: A profile for TV

Thanks for sending this out, Giuseppe. I would like to work on this effort.

I agree that we need to first discuss which meaning(s) of "profile" we wish to develop.

Here are two contrasting kinds of profiles to help understand the issues:

- Superset Profile
The HTML5 spec leaves many features optional that are important for interoperability (e.g. choice of XHTML and/or HTML, JavaScript, CSS, image types, media types, etc.) A Superset Profile would define, for example, that a range of devices would all support the full HTML5 spec, and additional provide requirements for many of the optional HTML5 features above. Every HTML5 browser makes decisions on optional features - the Superset Profile would just publish such decisions in a product-independent manner. This would increase interoperability amongst this set of devices and still be fully HTML5 compatible. Content written for the Superset Profile may not run on all HTML5-compliant browsers (if the content depends on optional features), but the Superset Profile would run all HTML5-compatible content.

- Subset Profile
There has been discussion at previous Web&TV Workshops that the full HTML5 spec is too expensive for some consumer devices. A Subset Profile would define, for example, a subset of HTML5 that uses fewer resources than the full spec. Content written for a Subset Profile would run on any HTML5 browser, but the Subset Profile device would not run all HTML5-compatible content.

There are a number of external organizations, particularly in the TV space, which have created HTML specifications which are Supersets, Subsets or both. Often, these external specs become obsolete when the W3C improves specifications, such as with HTML5, since the W3C has little or no knowledge of these external specs. Defining a TV Profile within W3C could provide a common profile spec for these external groups to reference. This would more closely align the external organizations with W3C and with each other. The W3C could also move these profiles forward on a regular basis to avoid obsolescence.

Thanks,
mav


On Dec 13, 2011, at 4:01 AM, Giuseppe Pascale wrote:

> Hi all,
> I believe is time to start to consolidate the experience we are gathering
> during our conversation in this IG. One way to achieve this could be to work together on a "Profile for TV".
> 
> As we discussed during our last F2F in Hollywood (notes available at [1]),
> there are different meaning that people usually associate to the word "profile" and for sure we will have to address this as a first step of our work and set out clear expectations.
> 
> With this email I would like to ask IG members:
> 
> 1) if they are interested to work on such a profile
> 2) what are your expectations for such a document.
> 
> about 1),to express you interest please send a reply to this mail to the
> list or privately to me. If there is enough support we will start a new TF
> working on this.
> 
> about 2), I'll list below my expectations, feel free to do the same in
> reply to this mail, adding your expectations or commenting on mine
> 
> 
> My expectations:
> 
> * The range of technologies available for web applications is so wide that an implementers necessarily need to make a choice on what to implement and when.
> 
> While in some ecosystems is fine to leave to each implementer to choose his own roadmap, in other ecosystems there is a need for coordination in order to harmonize the development cycle of the different stakeholders (CE manufacturers, Content providers, Content Authors, etc.) and provide a good user experience.
> 
> A profile can then also help on this aspect.
> 
> * The HTML5 spec, in some places, defers to other specs the role of defining
> how a feature would be mapped on other layers of the media stack.
> For example, for in-band tracks you can read
> "Set the new text track's kind, label, and language based on the semantics
> of the relevant data, as defined by the relevant specification."
> 
> While this is perfectly fine in the HTML5 scope, is clear that a mapping
> document (the "relevant specification") needs to be defined by someone
> depending on the infrastructure they are using. This discussion is
> currently going on in the Media Pipeline Task Force, but we eventually
> need to write down this mapping spec (as already suggested by Clarke at TPAC).
> 
> There are other areas where a similar work is needed (e.g. exposing metadata)
> 
> A TV profile could serve this purpose as well.
> 
> * Task Forces of the IG are trying to identify gaps in web technologies.
> Sometimes the outcome of the discussion is just that there is already a
> way to cover a use case with existing technologies. If this outcome is not
> documented somewhere, though, there is a risk that other people will have
> the same discussion again. Would be than good to document how certain use
> cases, relevant for the TV industry, can be covered with existing
> technologies.
> 
> A TV profile can help to document this.
> 
> * There are some identified gaps which are being addressed by work in one
> or more W3C WGs. Is good to document this activity so that people
> looking for that know where to find it.
> 
> This can also be documented in a profile.
> 
> 
> 
> [1] http://www.w3.org/2011/09/21-22-webtv-f2f-minutes.html#day1-am2
> [2] http://dev.w3.org/html5/spec/introduction.html#scope
> [3] http://dev.w3.org/html5/spec/media-elements.html#sourcing-in-band-text-tracks
> 
> -- 
> Giuseppe Pascale
> TV & Connected Devices
> Opera Software
>

Received on Wednesday, 14 December 2011 04:10:21 UTC