Re: minutes for HTML WG f2f, 2011-11-04, part 1

On 08/11/2011, at 10:58 PM, Peter Winnberg wrote:

> 2011/11/7 Cameron Heavon-Jones <>:
>> It's the difference between asking what it means to have an index of all <span> on a page over what it means to have an index of all <data> on a page.
>> You can turn <span> into a data container, but it will inherit a lot of meaningless data.
> First of all, the span element means that there is something here but
> no information about what it is. This is exactly the same as the data
> element as far as I can see.

Some discussion on the <data> element has indicated a potential for greater integration within html, if this were the case it would specify functionality over the element which would distinguish it from the generic <span> or <div> elements. There is also the possibility that some inference could be made over an non-attributed <data> elements so that this could still mean something: <data>Thursday</data>

The question seems to be whether data markup has a need and role within html or if it should be left to external technologies. The inclusion of a <time> element indicates that there is a need outside some data markup solutions for this kind of elementary data to be defined.

There is the question of what the nature of html is and is not. To break it down a little, hypertext markup is author-driven annotations over plain text to aid in machine understanding of the nature of the text (as a document at first, or otherwise). It would seem that if we have to define the start and end points of paragraphs, abbreviations, headers, footers, etc it would also seem that the language should also provide the ability for authors to declare which time they are talking about, or which location or any other value, aiding entity extraction. 


Received on Wednesday, 9 November 2011 13:16:46 UTC