W3C home > Mailing lists > Public > public-html@w3.org > December 2009

Re: ISSUE-76: If we fixed namespaces, does RDFa still have problems?

From: Benjamin Nowack <bnowack@semsol.com>
Date: Mon, 14 Dec 2009 15:06:50 +0100
To: public-html@w3.org
Message-ID: <PM-GA.20091214150650.962C0.1.1D@semsol.com>

On Fri, 11 Dec 2009 01:27:22 Manu Sporny wrote:

>RDFa 1.1 will allow for the use of RDFa without declaring any namespaces
>and using vocabularies and keywords instead. For example:
><p vocab="http://example.org/foaf.html" about="#robert" typeof="Person">
>   My name is <span property="name">Robert Ennals</span>.

Wow, how are stand-alone parsers supposed to translate this into 
RDF namespaces w/o de-referencing the http://example.org/foaf.html?

>> When the namespace problem is taken away, what differences are still
>> problematic?
>Any others that I missed?

IMO one of RDFa's biggest problems WRT to HTML authors is its RDF 
legacy. The attribute names are rather inconsistent and based on
one of the RDF community's worst specs, RDF/XML, which, too, has
too many language constructs for a flat learning curve. While
"about", "resource", or "property" sound familiar to RDF/XML 
authors (if there are any left, dunno ;), RDFa in general has the 
air of being "by RDFers, for RDFers". I do have my problems with 
Microdata as well (I would have created a syntax along 
 @item=[optional item id]
but it's still natural to write and code for, especially when 
JavaScript is involved. I'm a brainwashed RDFer, and still prefer 
Microdata over RDFa  (If only HTML validators wouldn't complain 
about the unknown attributes...).

On one of the other W3 mailing lists, someone did a test around 
the way existing RDFa parsers convert markup and how claimed 
"adopters" such as Google apply/support RDFa. It seems we're 
still far from a common understanding of the syntax and processing
model, this might also be evidence for the necessity for a simpler

Another issue is the lack of restrictions when it comes to adding
RDFa attributes to HTML nodes. In the attempt to keep compact 
markup, people add combinations of @typeof, @rel, and @property
on a single node and wonder about the extracted RDF (as seen in 
this thread) as added attributes may change the former graph 
model through implicit subject insertion.

The funny thing is, RDFa 1.1 (minus hacks such as @vocab) would 
only have to change some of its attribute names to become a 
proper microdata superset. If RDFa 1.1 requires new parsing 
libraries anyway, why not base it on Microdata? This could end 
the split-it-out discussion immediately and we wouldn't see 
this really disturbing situation of RDFers effectively voting 
*against* native RDF support in HTML. After all these years of 
uphill battles, it's a bit sad to see the small RDFa group harm 
the broader SemWeb community due to short-sighted motivations.

Related: I've seen several comments about lack of implementation
support for Microdata, so let me add mine:
* I've recently switched from RDFa (and home-grown poshRDF) to 
  Microdata in an open-source CMS that's going to be released 
* ARC[1] (a widely used RDF toolkit for LAMP systems) is 
  going to get Microdata support in one of the next releases.
  (I'll keep the existing RDFa parser in the bundle for the
  time being, though). Drupal's RDF modules are already using 
  certain components of ARC, so it's not unlikely that Drupal
  will get Microdata support in the future, too.
* FanHubz[2] is one example of a site where we used Microdata
  to embed global resource identifiers. The app is still 
  based on the earlier Microdata draft, but I don't think the
  upgrade to the latest spec will cause issues. So far, it's
  been very simple to process Microdata with JavaScript, and
  not having to inject prefixes in UIs that consist of nested 
  template snippets is very helpful.

Sorry for the possibly unfriendly tone here and there, but
I really think RDFa could leverage Microdata and should use 
this rare opportunity of getting at least a subset of the 
syntactical requirements directly into the HTML spec. Esp.
if the spec is going to be changed anyway.


[1] http://arc.semsol.org/
[2] http://fanhu.bz/

Benjamin Nowack

>-- manu
>Manu Sporny (skype: msporny, twitter: manusporny)
>President/CEO - Digital Bazaar, Inc.
>blog: Bitmunk 3.2 Launched - The Legal P2P Music Network
Received on Monday, 14 December 2009 14:07:22 UTC

This archive was generated by hypermail 2.3.1 : Thursday, 29 October 2015 10:15:55 UTC