Re: Telecon Agenda - Thursday 22 October 2009, 1500 UTC

Mark Birbeck wrote:
> [...]
> Some background:
> 
> In my view, Microdata is a political proposal, cobbled together over a
> weekend, for reasons best known to the author. (I have my own
> theories, but that's not important.)

The design is simple enough that I don't imagine it took long to 
physically write the text for the initial proposal (and it only took a 
morning to read and understand it and write an implementation), but it 
was preceded by months of discussions of embedded structured data use 
cases and issues with RDFa (e.g. lots of messages in 
<http://lists.whatwg.org/htdig.cgi/whatwg-whatwg.org/2009-January/thread.html>) 
and followed by feedback from many individuals and a usability study 
(<http://blog.whatwg.org/usability-testing-html5>), so it seems 
factually inaccurate to view it as being developed "over a weekend".

> However, what if instead of using his editor powers to add whatever he
> liked to HTML5, Hixie had proposed some changes to RDFa.

<http://lists.whatwg.org/pipermail/whatwg-whatwg.org/2009-May/019681.html> 
explains it as a series of changes to RDFa - stop using prefixes, drop 
compatibility with existing RDFa content and processors, rename the 
attributes to be more consistent, remove features that are more powerful 
than are needed for the given use cases. So it seems like it shouldn't 
be hard to interpret as a series of proposals, and discuss which ones 
are good ideas that should influence RDFa itself, and which are bad and 
should be rejected.

> [...] in Microdata you can express relationships using full
> URIs; what if Hixie had proposed to this group that we support that? I
> for one would have agreed with him. I know Steven would too, since he
> has mentioned the consistency of URIs and CURIEs in all attributes, a
> few times in the past, long before Microdata was proposed.

That idea has been discussed a bit between RDFa and WHATWG members long 
before microdata, e.g. in 
<http://lists.whatwg.org/pipermail/whatwg-whatwg.org/2009-January/018278.html>, 
but (if I remember correctly) there seemed to be little interest in 
actually updating the RDFa spec to support new features. If there's more 
interest now in updating the language, I'd agree it sounds like a good 
feature to consider.

> As to where we should do this; I think we should do it in RDFa+HTML.
> We should put some effort into a backwards-compatible solution for
> 'URIs everywhere', and so what if RDFa in HTML 'leapfrogs' RDFa in
> XHTML?

Seems good to design and publish any changes as soon as possible, else 
the amount of legacy content and tools will increase over time and it 
will become more painful to make changes.

> Regards,
> 
> Mark

-- 
Philip Taylor
pjt47@cam.ac.uk

Received on Thursday, 22 October 2009 14:35:06 UTC