W3C home > Mailing lists > Public > public-rdf-in-xhtml-tf@w3.org > June 2008

Re: geo location tutorial cut

From: Mark Birbeck <mark.birbeck@webbackplane.com>
Date: Mon, 30 Jun 2008 10:36:43 +0100
Message-ID: <ed77aa9f0806300236h366d30f6xc4c2f0b5857540dd@mail.gmail.com>
To: "Shane McCarron" <shane@aptest.com>
Cc: "Hausenblas, Michael" <michael.hausenblas@joanneum.at>, public-rdf-in-xhtml-tf.w3.org <public-rdf-in-xhtml-tf@w3.org>

Hi Shane,

> Okay - I think I understand the distinction.  And to some people that
> distinction might even be meaningful.  But what about joe-myspacer?  All he
> wants to do is say joe knows some_band.  I am certain joe is going to say:
>
>
>   myspace:joe foaf:knows myspace:some_band .
>
>
> is that wrong?  I dont think it is.  It is misleading, since what he is
> really saying litereally is "joe's myspace page knows some_band's myspace
> page".

This is the information resource/resource question, which Steven also
raised recently.

If myspace:some_band is *not* a web-page, then there is nothing wrong
with this. If it *is* a web-page, then it is bad practice because you
would now have an identifier that identifies two different types of
things (as Michael said). Even then, as far as RDF is concerned it's a
grey area. There is nothing to stop me saying that a camel is also a
fish, so I could say that the band and web-site are one and the same.

But I guess it's obvious that this will soon get messy; how can a
semantic database work if items can be two things at the same time?

You might argue that a URI isn't even one thing...so how can it be two
things! In your statement that A knows B, there is nothing to say that
A and B are web-pages, so who is know that there is anything going
wrong?

But the 'web-page-class' is such a common class of things to talk
about, that most people agree that we should treat them with care. So
it's generally accepted by convention that if something looks like a
web-page URI, then it probably *is* one. (We should also recall that
RDF was originally invented to talk about web-pages, and only later
became used for talking about camels, fish and bands; so in many ways
it's true to say that 'a URI is a web-page unless something tells us
otherwise'.)


>  But to the great unwashed - our target audience - is that
> distinction even meaningful?  Heck, some of those people might think their
> myspace page *is* their identity!

The problem is this: although you are right that an ordinary user is
not interested in the distinction, that's really irrelevant at this
level of the discussion. It would be like saying that people who use
washing-machines are not interested in whether electricity flows from
positive to negative, or vice versa.

All very true...but not a basis on which to build a science.

This need for something usable at the same time as providing something
that is solidly scientific and consistent, has caused RDF a lot of
problems in the past. People see the interminable discussions about
information resources, and simply assume that RDF is 'theoretical'.
But this theoretical distinction is necessary, as Michael pointed out;
regardless of what end-users want, you could not have an RDF database
that didn't know whether an item was a web-site or a monument.


>  But if it is meaningful, how can we help
> our audience to appreciate the distinction?  And more importantly, how can
> we help our audience to use the *correct* subjects and objects?

It's tricky... :)

Solving the distinction between the two resource types so that it is
easy for end-users won't be done by changing RDF, or ignoring the
distinction; it will have to be done in a way that makes it easier for
end-users to create the 'easy triples' whilst still allowing the
distinction to be made for the more 'unusual' ones.

In your example that essentially means making it easier to write about
bands than web-pages.

One possibility is that you don't say:

  I like the band 'Some Band'

but you say:

  I like the band with the web-page at myspace:some_band

The 'like the band with this web-page' property might be defined by
MySpace or by someone else, and authors wouldn't really need to know
the subtleties of why it's done--they would really just be saying 'I
like this band'.

Another approach is to simply say that every URL that looks like a
web-page but is not one, should end in '#', which is a well
established convention. Again, authors wouldn't really need to know
what's going on, and their editors/tools/whatever would just create
this:

  myspace:joe foaf:knows myspace:some_band# .

Of course, nothing is ever as easy as it looks, and if the band moves
their web-page, this falls down.

Which is why this whole question in RDF requires precision.

Anyway...I'm not saying any of these are the answer, but it shows that
we could find ways through this if we wanted to put some effort into
it.

Regards,

Mark

-- 
Mark Birbeck, webBackplane

mark.birbeck@webBackplane.com

http://webBackplane.com/mark-birbeck

webBackplane is a trading name of Backplane Ltd. (company number
05972288, registered office: 2nd Floor, 69/85 Tabernacle Street,
London, EC2A 4RR)
Received on Monday, 30 June 2008 09:37:24 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 30 June 2008 09:37:25 GMT