Wednesday, March 05, 2008

Mashups, microformats and the mobile web Panel


That's me on the left, standing. Then comes Conor, Ben and Bill.

The panel was good fun. The big takeaway for me was the amazing amount of consensus that microformats are the way to go. Two years ago, I was getting shot down in flames for suggesting that the future of semantic markup lies *inside* existing schemas. Times have changed.

10 comments:

Noah Slater said...

Not everyone agrees with that. Have you looked at the W3C's GRDDL effort?

Sean said...

Noah,
Yes, I have looked at GRDDL.
Sean

Kevin Feeney said...

Hi Sean,

I Enjoyed chatting with you on Saturday night - thanks for giving me the time.

I suspect that you would get a very different reaction to microformat arguments in different venues. Bloggers and social-media types are very wedded to the simplest way of making stuff work - NOW! I can think of a few places where you would get a properly horrified reaction.

Kevin Feeney

about said...

Heh, nice response. Okay, I will elaborate... where/why do/would you feel that microformats outshine GRDDL?

Sean said...

@about,
GRRDL is about RDF extraction. Fine. Its very useful. uFormats provide a basis for overlaying a custom directed acyclic graph on top of the DAG provided by the underlying XML schema (XHTML, ODF etc.). Its very useful for that.

There is more to the world than RDF triples in the same way that there is more to the world than superstrings. Just because everything in the world can be thought of as bundles of superstrings does not mean that it is the appropriate level of either representation or discussion :-)

Sean

about said...

But GRDDL does exactly the same as that, allowing you to use custom class names and profile specific markup to convey meaning and semantics.

Sean said...

@about,

I suspect we are missing each other at 30,000 feet.

Not all structure content sensibly reduces to triples. That is why the world needs XHTML, ODF etc. as well as something like RDF or (for us old-timers) Prolog at the back-end.

Sean

pkeane said...

Here's a nice & apropos posting about RDFa vs. microformats:

http://www.snee.com/bobdc.blo/2008/02/the_future_of_rdfa.html

And while microformats might fill some nice niches, it seems like the future rests more w/ RDFa.

--peter keane

Sean said...

Peter,

I hear ya and I'm familiar with Bob's blog but I'm going to stick to my guns on this one. Microformats hit a real good 80/20 point. RDF may be intellectually all-encompassing but that is, generally, not a strong indicator of mass market adoption. Beware the

uber-abstraction/

Microformats have their problems for sure. Accessibility concerns, name-clash concerns etc. I predict that these will be addressed in the somewhat messy fashion that most "bugs" are addressed in sucessful technologies. I very much doubt that we will all wake up one day and decide to start from a completely different n+1 levels-of-abstraction starting point.

Sean

pkeane said...

100% agreed on the dangers of uber-abstraction, but I'd contend that RDFa (or let's just say the opportunity to declare what I "mean" when I embed this bit of structured mark-up) presents a lower barrier to entry than Microformats, since I can essentially declare my *own" microformat which might or might not be adopted by others, but meanwhile offers interop (if only w/ in a closed system) opportunities. I see RDFa as "RDF for the masses". I'd love to see the Microformats efforts being more attuned to RDFa and working towards "registering" particular constructs while not disallowing others. Kind of like the IANA registry of Atom link relations -- I can use one of those, or declare my own w/ a URI.