[microformats-discuss] The adoption of syndication feeds -> microformats

Danny Ayers danny.ayers at gmail.com
Thu Oct 6 05:52:49 PDT 2005

On 10/6/05, David Janes -- BlogMatrix <davidjanes at blogmatrix.com> wrote:
> I just had a little nap; upon waking up I feel I may have been a little
> unfair to the FOAF crowd. The reason I used it as an example is that
> I've seen so many broken FOAF files it seemed evident that even
> supporters weren't particularly interested in keeping the files up to
> date (implying that there aren't that many consumers).

Yep, FOAF is actually an *awful* example for the (valid) point you
were making ;-)

FOAF primarily provides a *model* for the Person construct, attributes
of it and relations between those constructs. FOAF files are one kind
of serialization of that model (RDF/XML). But much of the same
information can be expressed in other formats, such as XFN. The
particular serialization is of secondary importance.

I think there's a nice analogy here: In the same way XHTML acts as a
substrate for microformats, RDF is the base for micromodels.

Microformats are being designed in such a way that the syntax of each
can be (relatively) freely mixed in a document. RDF vocabularies are
designed in such a way that the model of each can be freely mixed on
the Web.

If there is a model and a mapping from the syntax to that model, it is
possible to view microformats as domain-specific serializations of RDF
- the general technique is the GRDDL [1] to which Brian referred).
It's possible to automatically read microformats into RDF systems -
the Redland RDF toolkit has support for this built in.

A little while back in an exchange with Tantek he said something along
the lines of "it doesn't really matter what kind of store you have in
the backend", in the sense that microformats could equally be
produced/consumed by SQL DBs, XML DBs or RDF DBs. True enough, he was
emphasizing the role microformats can play in exposing a standard
interface. But in the RDF case this can go further - a shared data
model can be exposed to the Web as well.

In some respects microformats and RDF are orthogonal, one being about
syntax the other about model. But they're both pointing in the same
general direction, improving the Web through explicit data. When
techniques like these are used to put unambiguous machine-readable
data on the Web, the lower-case and upper-case semantic webs are one
and the same.


[1] http://www.w3.org/TeamSubmission/grddl/


More information about the microformats-discuss mailing list