W3C home > Mailing lists > Public > www-tag@w3.org > April 2006

RE: RDDL as a microformat?

From: David Orchard <dorchard@bea.com>
Date: Tue, 18 Apr 2006 13:05:25 -0700
Message-ID: <E16EB59B8AEDF445B644617E3C1B3C9C016D8BCF@repbex01.amer.bea.com>
To: "Robin Berjon" <robin.berjon@expway.fr>, <www-tag@w3.org>

Hi Robin,

Like many other folks, I've been interested in mformats.  I'm not sure
about the utility of RDDL as a mformat, though it would be easy enough
to write up.  BTW, I think your proposal is a little more than a mformat
for Xlink than for RDDL.  I think the big upside to mformats is
embedding structured information in a lot of web pages, like
reviews/cards/events, etc. and then having this information extracted.
I don't think that are that many RDDL pages that are out that there that
would be extracted for searching nor are the queries that interesting.  

I see a big use case in software engines like Technorati or kritx or
whatever extracting and then doing something with the info.  For
example, the scenario I'd love to see is a google/yahoo/msn map that
shows all the 4/5 or 5/5 rated restaurants in Vancouver .  

Now I can't see the equivalent use case for RDDL.  In my mind, the main
use case for RDDL is finding information about a given namespace name.
That is, the namespace name is a specific thing in hand.  Now imagine an
engine that grabs RDDL docs.  I don't think there's much utility in
asking questions like "find me all the namespaces that have RelaxNG
schemas associated?".  

Cheers,
Dave 

> -----Original Message-----
> From: www-tag-request@w3.org [mailto:www-tag-request@w3.org] On Behalf
Of
> Robin Berjon
> Sent: Monday, April 17, 2006 4:36 PM
> To: www-tag@w3.org
> Subject: RDDL as a microformat?
> 
> 
> Hi,
> 
> while mulling over the current state of RDDL, which doesn't seem to
> have changed a lot recently (unless I missed something), I wondered
> if instead of XLink or the RDF variants proposed in its stead it
> wouldn't be simpler to just use a microformat. I'm not convinced it's
> a good idea, but I thought I'd dump it here in case anyone would be
> interested. From the RDF extraction point of view it's reasonably
> similar, but from the author's side it's much more friendly than
> XLink attributes that don't have the names of what they do. It also
> makes all the links directly usable by humans, without duplicating
> the metadata (as the RDDL example below does, using the same link
> twice).
> 
> Instead of:
> 
>    <rddl:resource xlink:title="RELAXNG Schema"
>                   xlink:arcrole="http://www.rddl.org/purposes#schema-
> validation"
>                   xlink:role="http://relaxng.org/ns/structure/1.0"
>                   xlink:href="foo.rng"
>                  >
>      <p>A RelaxNG Schema <a href="foo.rng">foo.rng</a> for FooML</p>
>    </rddl:resource>
> 
> one would use:
> 
> ...
>   <head profile='http://www.w3.org/2006/04/rddl'>
> ...
>    <div class='rddl' title='The RelaxNG schema for FooML'>
>      If you want to validate
>      <a class='purpose' href='http://www.rddl.org/purposes#schema-
> validation'>validate</a>
>      you can use the
>      <a class='nature' href='http://relaxng.org/ns/structure/
> 1.0'>RelaxNG schema</a>
>      found at
>      <a class='locator' href='foo.rng'>foo.rng</a>
>    </div>
> 
> --
> Robin Berjon
>     Senior Research Scientist
>     Expway, http://expway.com/
> 
> 
Received on Tuesday, 18 April 2006 20:06:00 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Thursday, 26 April 2012 12:47:39 GMT