Re: Display values for RDFa object URLs

Hi Ivan,

> I swear I did not do it on purpose!

He he...I believe you. ;)


> But I just made a page where I've put:
>
> [[[
> package is also <a
> href="http://dev.w3.org/2004/PythonLib-IH/pyRdfa/">available for
> download</a>.
> ]]]
>
> Yes, this is not RDFa but, had I wanted, I could have said something like
>
> [[[
> <a rel="namesp:code"
> href="http://dev.w3.org/2004/PythonLib-IH/pyRdfa/">available for
> download</a>
> ]]]
>
> The text of the link is, sort of, o.k. for the flow of the HTML text,
> but I do not think it is o.k. for the label of
> http://dev.w3.org/2004/PythonLib-IH/pyRdfa/.

Sure...but rdfs:label is not *the* label for the resource, but *a*
label. It's not a big deal...it's not like I'm suggesting making it
dc:title or something.


> However, if I take your
> approach, _I do not have a choice_, that _will_ be the rdfs:label. And
> that is what I do not like: I've lost control.

Not at all. You've got an extra triple which you can use if you like.


> By the way, one of the differences in our perceptions may come from your
> remark:
>
> [[[
> we try to provide an *interpretation* of the mark-up in triple form,
> ]]]
>
> which is not the way I look at it. We do not try to interpret the
> mark-up; we give tools to the user to add 'metadata' (yey, I used the
> m-word:-) to his/her own text in a specific mark-up.

I didn't say we don't add metadata.

My point has always been that HTML already comes with a lot of
metadata features and one of the things we're doing with RDFa is
formalising the interpretation of those. Some of this can be done
quite precisely since the HTML spec itself defines things sufficiently
clearly. A good example of that would be:

  <link rel="next" href="o" />

With this example, although HTML doesn't mention RDF, it gives us
enough information to say that we can interpret this as:

  <> xh:next <o> .

Now, I realise that since HTML wasn't defined with RDF in mind, there
are lots of situations where the 'interpretation' of HTML's metadata
capabilities is more art than science. But I'm not sure that the
creation of one or more rdfs:label's is one of them.

The RDFS definition of rdfs:label is merely "a human-readable version
of a resource's name". It's not "the one and only", so it doesn't seem
a big deal to me to add text from inside <a> to our list of triples. I
actually did this in my parser for testing purposes quite a long time
ago, since without it, there were certain things that couldn't be done
with links gleaned from a page (such as 'add this link to
del.icio.us'). I'm not saying you have to use the label. :)


> I guess we can safely agree that we disagree on that:-) and let the
> group vote on this...

Of course. ;)

Regards,

Mark

-- 
  Mark Birbeck, formsPlayer

  mark.birbeck@formsPlayer.com | +44 (0) 20 7689 9232
  http://www.formsPlayer.com | http://internet-apps.blogspot.com

  standards. innovation.

Received on Friday, 10 August 2007 13:00:04 UTC