Adding support for custom variants of HTML is not really in line with the goals 
of the WebKit project. It sounds like CE-HTML and OIPF are both HTML variants. 
I don't think these types of features are a good fit for WebKit, because:

(1) Custom HTML dialects tend to create excessive maintenance burden for the 
contributors who are not interested in the feature. WAP is a historical 
example. We much prefer features that either have broad mainstream interest, or 
at the very least can be implemented with minimal intrusion into core code.

(2) Most contributors to the WebKit project strongly believe in a "One Web" 
vision, where the same full core specifications are used in all the contexts 
where Web technology is useful - no special dialects for mobile, or tv, or 
whatever, it's all one web. Custom HTML dialects go against that vision, so the 
value of adding support would have to be very high to even consider it.

I strongly recommend that the HbbTV effort should use HTML5 and other Web 
technologies directly, rather than profiling and extending the Web.

Regards,
Maciej

On Oct 8, 2012, at 7:11 AM, Mark Toller <mark.tol...@samsung.com> wrote:

> Hi,
>  
> I'd like to ask the Webkit developers their opinion on providing some support 
> for HbbTV [1] within Webkit. Hybrid Broadcast Broadband TV or "HbbTV", is a 
> major new pan-European initiative aimed at harmonising the broadcast and 
> broadband delivery of entertainment to the end consumer through connected TVs 
> and set-top boxes.  The HbbTV standard is proving to be very popular, TVs and 
> STBs supporting HbbTV are shipping in huge numbers throughout Europe.  HbbTV 
> is built on top of OIPF [2], which in turn is based on portions of CE-HTML 
> [3].
>  
> Our lab, Samsung Electronics Research Institute (SERI), has been heavily 
> involved in HbbTV and our current solution is based on Webkit. We would like 
> to provide our changes back to the community.
>  
> I know that support requests for CE-HTML have been briefly touched upon in 
> the past. As I understand it, the main objection to providing support within 
> WebKit is that the CE-HTML specification is not freely available, and thus 
> restricts the number of developers who can fully understand it and therefore 
> provide fixes / support.
>  
> In reality, much of the CE-HTML specification simply profiles which parts of 
> the W3C standard behaviour are mandatory, optional and/or recommended. OIPF 
> then profiles CE-HTML (dropping some requirements, extending others to match 
> W3C/HTML5), HbbTV profiles out even more of CE-HTML.
>  
> Other parts of OIPF and CE-HTML do not need to be implemented within Webkit 
> itself. Some can be implemented as object plugins (e.g. AV Control and local 
> video), while others, such as the JavaScript classes required, can be 
> inserted into the JavaScriptCore at runtime.
>  
> What I propose is to provide the basic support required within Webkit in 
> order to at least load the XHTML portions of HbbTV applications and provide 
> the correct key handling to drive them. In order to provide 'full' HbbTV 
> support, implementations would need to provide the plugins and additional 
> JavaScript classes to complete the picture.
>  
> For instance, by simply adding support for the document mime type handling of 
> application/vnd.hbbtv.xhtml+xml and application/ce-html+xml, many HbbTV 
> applications will load and display the main page, and several will also 
> correctly navigate around the application correctly.
>  
> Regards,
>  
> Mark.
>  
> [1] Hybrid Broadcast Broadband TV - http://www.hbbtv.org/
> [2] Open IPTV Forum - http://www.oipf.tv/
> [3] CEA, CEA-2014-A, Web-based Protocol Framework for Remote User Interface 
> on UPnP Networks and the Internet (Web4CE)
>  
>  
> _______________________________________________
> webkit-dev mailing list
> webkit-dev@lists.webkit.org
> http://lists.webkit.org/mailman/listinfo/webkit-dev

_______________________________________________
webkit-dev mailing list
webkit-dev@lists.webkit.org
http://lists.webkit.org/mailman/listinfo/webkit-dev

Reply via email to