Discussion with Ian and Henri about HTML5+RDFa (part 1/2)

I spent some time last week on the phone with Ian Hickson (Editor of the
HTML5 spec) and some time today with Henri Sivonen (Mozilla Foundation's
representative for the HTMLWG and lead developer on the HTML5 validation
suite).

The goal of the phone calls were to help clear up some misconceptions
that I believed had been brewing in both work groups for a while. What
follows is a summary of the discussions. I don't necessarily agree with
anything below, but am passing it on so that there is a better
understanding of what the WHATWG is requesting and the major issues that
they see with RDFa as it stands right now.

The discussion with both Ian and Henri was very pleasant, friendly, and
rational - both were able to clearly outline the current issues that
they saw with RDFa and proposed a number of fairly straight-forward ways
of moving the RDFa in HTML5 discussion forward. I trust that both of
them will correct any mistakes that I make in conveying the gist of what
they said.

Discussion with Ian Hickson
---------------------------

Ian's primary message was that the RDFa proponents have not clearly
outlined a list of use cases and why such use cases should be supported
in HTML5. He is concerned that without this basic information, which is
required for most new HTML5 features, it becomes difficult for him to
argue that there is an empirical basis for placing RDFa into HTML5.

I have started to gather the list of use cases that started RDFa as well
as the list of use cases that have sprung up around RDFa over the past
several months (add more if you know of cases in the wild):

http://www.rdfa.info/wiki/rdfa-use-cases

Some of Ian's other issues included:

- He was concerned that a generalized solution such as RDF can always be
solved with a much more specific vocabulary and mark-up solution (such
as HTML5 or Microformats) and that the more specific mechanism for
semantics expression should be favored and supported. For example, there
is an <article> tag in HTML5, which is favored because it expresses
semantics and authors would probably be more prone to using it because
they can associate CSS with the element, where they cannot do the same
with the RDFa attributes. The same would apply for the <audio> and
<video> elements in HTML5. I believe that would prefer to focus on
allowing a markup mechanism that supported RDF as well as other
semantics expression mechanisms - IF, it can be shown that it is
desirable for authors to embed semantics in web pages.

- If a decision had to be made today, Ian believes that it would go to a
vote which would be dangerous for both WHATWG and RDFa because the
outcome would effectively be random. Nobody really knows what the 300+
members of WHATWG, ~60 of which vote on a regular basis, think about
RDFa. It would be safe to assume that a significant portion do not know
enough about RDFa to vote on it.

- He re-iterated that he does not have anything major against RDF or the
idea of semantics expression in HTML5. He does have certain issues with
RDFa, but would like to see the use cases that we worked from in order
to better understand the problem domain. He seemed sincere on working
towards a solution of some kind.

The discussion with Henri Sivonen will be in the next e-mail.

-- manu

-- 
Manu Sporny
President/CEO - Digital Bazaar, Inc.
blog: Bitmunk 3.1 Website Launch
http://blog.digitalbazaar.com/2009/01/16/bitmunk-3-1-website-launch

Received on Monday, 19 January 2009 23:06:40 UTC