[Odrl-version2] Microdata and RDFa HTML4/5 specs

[Steven Rowat]

Greetings,

Summary of this email:

In brief, I wish to know what to do about the RDFa/Microdata split: as 
of today, August 18/09, there appears to be a confusing fork between 
Microdata and RDFa; and no sign of a resolution; so that html writers, 
such as myself, will be left wondering which to use. This is 
particularly troubling after over a decade of hoping for a widely 
accepted universal metadata standard. So, in the light of repeated 
calls for use-cases to help decide which of Microdata/RDFa is better 
and why (see links below), I'm hoping that some progress has been made 
in this that can be told to me; or, failing that, even that 
considering my own use-case might be a help, however small, in 
bringing that decision about.

Details:

For the purposes of this email I might describe myself as an 'end-user 
of extended metadata'. In more common language, I'm designing a 
complex web site with data that is not easily cataloged by search 
text-based engines -- audio recordings, graphics, and metaphor-based 
poems whose themes are not fully or accurately contained in their own 
literal text. Not only do I wish these things cataloged accurately for 
software-based searches, I wish them available for DRM (including 
copyright statements and rights-use prohibitions and permissions, as 
well as, eventually, prices and protocols for direct sales, through 
the browser).  These protocols are, for instance, provided by ODRL 
(the Open Digital Rights Language), which already has draft protocols 
for both microformats and RDFa implementation (see links on the 
www.odrl.net page).

All well and good - then I discovered the new Microdata spec 
introduced into HTML5 that, according to many observers replaces RDFa, 
and, according to some observers at least, for no good reason. This 
was a very confusing discovery. More, when I went further into the web 
discussions about this replacement I realized it has been the subject 
of a heavy, active, and ongoing discussion, continuing from May 09, 
right up into this week in August.

And today, August 18, Draft HTML5 specs have appeared with today's 
date -- and only Microdata; no mention of RDFa is found at all in the 
document (though RDF is mentioned, as one of several protocols 
Microdata can be exported to.)

I've put a list below of some of the discussions around this issue. 
They are extensive, and I'm sure I could find many more.

So, my question: what's a guy to do? :-) . Or, more importantly, what 
are the millions of other unique content producers who write html all 
over the world going to do now?

    a) Learn RDFa and external vocabularies, like ODRL, for it (and 
just ignore Microdata...do I need it? Maybe it will go away)?
    b) Learn to code Microdata, ignoring RDFa (and if so, how do 
external protocols like ODRL fit with Microdata? Will there be a new 
ODRL protocol required for Microdata?)
    c) learn all of Microdata, RDFa, and the external (like ORDL) 
vocabulary for RDFa, and code my pages so that they have Microdata 
that calls RDFa that calls the external vocabulary?
    d) wait and do nothing, and hope either Microdata or RDFa 'wins' 
on the web? (Note: by my reckoning, it's not just *years* since a 
widely-accepted extensible metadata protocol was expected on the web; 
we're into *decades*)

Any suggestions as to which of these four might be preferable at this 
time, and why?

At this point, I'm certainly leaning towards *something* being 
adopted; it seems like a major, even show-stopper, issue. Although 
conceivably one that can only be resolved at a different level of the 
decision-making structure of the web than the one where it is 
currently stuck.


Best regards,


Steven Rowat



Here are some of the links I investigated about the Microdata/RDFa split:

Editor's draft 18 August 09 of the HTML 5 Vocabulary:
http://dev.w3.org/html5/spec/Overview.html#concept-item

FAQ from the W3C "Frequently Asked Questions (FAQ) about the future of 
XHTML"-- see #6 about extensibility, and #7 especially: "What are 
W3C's plans for RDFa?"
http://www.w3.org/2009/06/xhtml-faq.html

ISSUE-41 Decentralized-extensibility
http://www.w3.org/html/wg/tracker/issues/41

Re: Does anyone like microdata? (W3C list). Has defenses for Microdata 
protocol in HTML5.
http://lists.w3.org/Archives/Public/public-html/2009Jun/0811.html

Re: Why bound prefixes are an anti-pattern in language design: (W3 
discussion list)
http://lists.w3.org/Archives/Public/public-rdf-in-xhtml-tf/2009Aug/0075.html

"Intertwingly" Microdata:
http://intertwingly.net/blog/2009/05/12/Microdata

Use Cases and Comparison of RDFa/HTML5 Microdata, Shelley Mon:
http://realtech.burningbird.net/semantic-web/semantic-web-issues-and-practices/use-cases-and-comparison-rdfahtml5-microdata

What You Can't Do with HTML5 Microdata, Jeni Tennison:
http://www.jenitennison.com/blog/node/103