Re: FaceBook taking over the web, and semantic web

(cc: list trimmed; one list is usually enough, two is plenty... three means
people start complaining)

On Fri, Apr 23, 2010 at 4:12 PM, ProjectParadigm-ICT-Program <
metadataportals@yahoo.com> wrote:

> Facebook plans to take over the web, by creating open graph based semantic
> content.
>
> See:
>
> http://blog.newsweek.com/blogs/techtonicshifts/archive/2010/04/22/facebook-f8-internet-open-social-graph-semantic-web-twitter.aspx
>
> FaceBook has come under fire for selling information about users to
> marketing companies.
>
> Is this a serious threat to the open structure of the Internet and the
> linked data paradigm as envisioned by the W3C?


There are plenty of things that people worry about re Facebook. I think
putting structured metadata in public Web pages needn't be one of those
things. We might wish the RDFa markup was tidier in syntax or semantics,
etc., but I don't think putting the director's name/ID into a
movie-describing page, or cuisine into a restaurant-describing page is a
cause for concern.

My main worry about Facebook is one shared with many of the other huge
profile-hosting and content-hosting sites: people are living their online
lives inside someone else's dns domain name. If Facebook provides were
http://johnsmith.example.name/ instead of http://facebook.com/johnsmith/ ie.
spread around and portable, I'd be much happier. So this new attempt to
reach out into the non-facebook.com Web is interesting. It strikes me as
much healthier if the main Web presence for a restaurant is at a Web site
owned/controlled by the restaurant, rather than on an internet domain owned
by Facebook (or Google, Apple, Microsoft, Yahoo, Twitter, whoever...). If
their new-found ability to make normal Web pages function like Facebook Fan
pages, then I think the trend over all is healthy. After all you can always
chop out the Facebook-specific markup in a few years, if you decide you
don't like them. Whereas if you go around putting
http://facebook.com/yourname or http://twitter.com/yourname as your main
URL, and encouraging everyone to link to it, then you are making it hard to
ever leave those sites.

Now when they say they're putting 'people at the heart of the Web', we know
that's a bit of spin and the goal is to put Facebook-mediated-people at the
heart of the Web. No suprise or shock there. But despite that, much of the
message in this week's F8 keynote is singing the Semantic Web song - about
connecting and mapping out a giant global graph linking all of us to the
things we make and do, from blogs and books to movies and restaurants. So -
well of course they want to control a big chunk of the Web, they're a big
tech company; to be upset of suprised at this is like being dissapointed
that foxes want to eat your chickens. Blame capitalism or something! Now
just because these mega-companies are naturally self-interested, it doesn't
mean everything they do is dumb, bad or dangerous. I think in this case, the
markup is at least fairly well separated. The meta tags are pretty neutral
and express something of what the page is about, while the LIKES iframe is
the bit that ties you into the Facebook world. Any company the size of
Facebook (or Google, or ...) is too big to think of as if it were a single
sentient entity. There are lots of different forces and trends in these
companies, and often plenty of people there who want to do the right thing
for the Web.

If their 'open graph' work lets more happen on user-owned dns domains rather
than in facebook.com, that strikes me as good. If they're doing this by
putting explicit metadata in pages using an RDFish object/property model,
... that's all for the good too. And we can be sure they have enough
competitors out there that people will soon enough realise that there's only
really one 'open graph' we can all rely on long-term, and that's the Web
itself.

Those who are really worried would be better placed building fabulous
opensource tools that exploit all this rich linked data we've gotten
published in RDF. I think that's the real advantage folk at Facebook have
over the rest of us. We can't easily slurp in and process data on such a
scale. Maybe there is scope for some collaborative projects in that
direction, at least using the public parts of the 'giant global graph'...

cheers,

Dan