Re: Now it's RDF vs Microformats

On 24/02/07, Paul Walsh, Segala <paulwalsh@segala.com> wrote:
>
> > -----Original Message-----
> > From: public-sweo-ig-request@w3.org [mailto:public-sweo-ig-request@w3.org]

comments like "The thing about
> > RDF is that no-one has yet demonstrated any real-world reason to care
> > about it" seem to beg for some SWEO intervention.

Absolutely.

> > I think Danny's response is well-worth reading if you haven't already,
> > though I remain personally unconvinced that the original post is
> > "well-argued".
>
> I think Danny was way too kind - the original argument doesn't stand up for
> me. In fact, it demonstrates a very confused/fuzzy opinion.
>
> Some opinions really are more equal than others - Danny's was more equal in
> this case :)

Maybe, but I'm pretty sure that if I'd started with strong
disagreement there would less likelihood of anything else I said being
considered with an open mind. What definitely is wrong with my
response is that it's long-winded and mostly handwaving.

While things like tutorials, toollists etc. are definitely needed
(they exist, but they're not very well collated), and wordy blog
comments may help a little, they don't in themselves offer anything
attractive to the typical web developer.

Benjamin's comment earlier was spot on:
[[
Try to avoid talking about "potential" uses, it's a "shake head and
laughter" trigger in MF circles.
]]

I really struggled when writing that comment to think of
demonstrations of "real-world reasons to care". There are a lot of
demos around that prove various invidual concepts (e.g. SPARQLing),
but their coolness only really appears when you extrapolate and join
the dots. It's not obvious.

I think Lee's right in saying the microformats community in general
aren't interested in being cooperative with the RDF crowd (and no
doubt some like the "we succeeded where RDF failed" rallying cry).
It's understandable, we haven't *demonstrated* there's anything we can
offer.

Personally I'd be inclined towards a "let them get on with it" kind of
approach, because they are oriented towards good practice with
standards, good for the web. Except firstly there are a few holes -
profile URIs being a big one, microformat data isn't GRDDLable without
them, but that's not a priority to most microformats folks. The other
question is what happens next. There is a clear evolutionary path from
domain-specific microformats to more generic data embedded in HTML.
We've even got two specs for it - eRDF and RDFa. But there's a very
good chance of other paths being taken, ones which aren't so
RDF-friendly.

If you go back the original syndication wars [1] between "simple" RSS
and RSS/RDF, the immediate outcome was two branches both getting
considerable adoption. But the only applications for RSS were simple
aggregators/newsreaders, and although ("simple") RSS 2.0 is a bit
rubbish, it's more-or-less adequate for that use. In terms of
deployment, I would guess "simple" is a now lot more widespread than
RSS/RDF, especially if you include the bugfixed version Atom. (There's
irony in that much of the motivation for Atom came from syntax issues
with RSS 2.0).

While RSS 1.0 was/is perfectly good RDF data, adoption and development
of RSS was largely driven by a single-purpose application, effectively
one domain: blog/newsreading. The net result is something optimised
for that application, the generalization offered by RSS was switched
off.

Microformats do cover a variety of domains, and can potentially target
a raft of different applications. But still there's likely to be a
tendency for the potential to be viewed in an unnecessarily narrow
fashion, and the evolution of microformats being influenced by this.
Sooner rather than later we could do with some more interesting apps
that consume and use microformat data.

Bleah, this has turned longwinded and mostly handwaving. Skip to the conclusion:

"Show don't tell".

Commenting on the post mentioned here would have been a lot easier and
more effective if I'd been able to point to a list of relevant,
compelling apps (and/or screencasts).

Cheers,
Danny.

[1] http://diveintomark.org/archives/2002/09/06/history_of_the_rss_fork

-- 

http://dannyayers.com

Received on Saturday, 24 February 2007 12:11:25 UTC