My suggestion would be that Drupal should generate a unique identifier
on the node itself (using @about) since this will then work in any
context.
Agreed! and that's the approach we've been using on the RDFa prototype
site at http://drupal.deri.ie/cheese/.
For example, if I search for items, then each node in the
search results would have its own @about value, and so be
self-contained. Similarly, if I use Views to create some new block or
something, by having @about on each node, everything 'just works'
within the block.
search results and node titles in blocks remain to be annotated and
that'll be one of the goals for the sprint next week. Thanks Mark for
your detailed replies!
Steph.
Mark Birbeck wrote:
Hi Stephane,
This brings me to the next topic! In RDF, the URI of a resource should be
different from the URI of the page that describes it. I've heard this is not
so important with RDFa since the RDF is self contained. Is that right?
I'm not sure what lies behind this point, but if I understand it
correctly, I'd have to disagree; ultimately the data is RDF, whether
carried via RDFa, N3, RDF/XML, or whatever. So we still need to solve
the same problems, and we still need to ensure that there is no
confusion between items.
Crafting a URI for a resource described at http://example.com/node/1 is easy
by appending #self, #me or #this, and this can be automated. However in the
particular case of a blog post for instance, the page and the resource it
describes are in fact the same thing. Shouldn't they have the same URI?
This is where it starts to get esoteric. :)
My suggestion would be that Drupal should generate a unique identifier
on the node itself (using @about) since this will then work in any
context. For example, if I search for items, then each node in the
search results would have its own @about value, and so be
self-contained. Similarly, if I use Views to create some new block or
something, by having @about on each node, everything 'just works'
within the block.
Of course, even in this scenario, you could possibly get away with
having the same URI for a blog post and the page that carries that
post, but I'd still advise against it.
By merging the two URIs you are effectively saying that you don't
envision a scenario where you would like to say things about a blog
post, independent of the page that carries the post...and as we know
in software, never say never.
A good example of where things can start to break down comes when
republishing blog posts (or YouTube videos, or Flickr images, etc.)
from other sites onto your own site; the subject of the post should
really refer to the original post's location, and not the URL on the
republishing site.
In short, I'd say that you can't go wrong if you keep the two
separate, whilst if you try to cut corners by merging the two URIs,
you could be storing up problems.
Regards,
Mark
