Re: Some Requirements

On Wed, 2003-09-24 at 07:13, Sandro Hawke wrote:
> On behalf of RDF authors everywhere, I'd like to make a URI which is
> usable in several, simultaneous ways: 
> 
>   r1.  As a not-overloaded name, essentially a logical constant term,
>        as specified in RDF Semantics [1].   The URI will be my name
>        for something; others may make other names for the thing or
>        reuse my name for it. 
> 
>   r2.  As a web page address for human-readable content, working in
>        currently-deployed browsers, giving users direct access to web
>        content which I supply associated with the URI;
> 
>   r3.  As a web address for RDF/XML content, allowing simple systems
>        to fetch a small to medium size knowledge base, which I supply,
>        associated with the URI
> 
>   r4.  As the address of a query answering service, allowing more
>        complex systems faster access to larger knowledge bases, which
>        I also supply associated with the URI

r1 and r3 are easy to meet with any ftp server. I don't
think r4 is a good architecture for query services.
I think query answering services should be related
to KBs by ordinary triples, not by URI munging.

Whether r2 can be met at the same time is unclear to me.
Do "currently-deployed browsers" support client-side XSLT?


> I can think of a few other uses, but if you can give me these four, i
> think I can figure out the rest.
> 
> Also, ideally:
> 
>    x1.  This would all work beyond the time over which I can
>         reasonably expect to control any particular domain name or
>         website;
> 
>    x2.  This would work even if a community of use arises which
>         disagrees with me about what the URI names.  I made it
>         up, I was the first to use it, no one should be able to
> 	make me into a liar by convincing people it meant something
>         different. 

I don't think that's at all feasible. You can't propagate
information for free; either you pay the price as publisher
(http/dns) or you get the world to support your views
(freenet, NNTP, etc.).

I don't think we should expect any technology to prevent
the winners from rewriting history, as it were.

My intuition says I should be able to make an argument
from information theory and economics about why you're
basically asking for a perpetual motion machine. But
I need to study information theory and economics some more ;-)

> For example, one of my little projects is to implement an OWL
> reasoner.  I call it Surnia, and I'd like to tell the world about it
> using all appriate means.  Basically that means I need a website for
> it (r2 above), but I also want to support RDF applications talking
> about it (r1) and learning about it (r3 and r4).
> 
> More specificly, the OWL Test Results page [2] needs an r1-style URI
> for Surnia so it can merge results from different sources, an r2-style
> URI so users can click to get more information, and r3- or r4- style
> URIs so it can show users data about it.  Actually, it doesn't *need*
> any of these and it doesn't really have any of these right now,
> because so far I haven't been able to make a URI like I want!  So
> instead there are three different URIs in use (no r4 yet), which is
> getting pretty clumsy.
> 
> About this point, settling for having three or four different URIs
> starts to look okay, but if I do that, I'm certainly not going to use
> an HTTP URI for r1 (it buys me nothing), ... so keep that in mind if
> you urge settling.

Using an HTTP URI for r1 buys you the ability to use the same
URI for r1 and r3... or at least: a syntactically related
uri, ala http://.../surnia-r3#surnia-r1 .


> As another example, I have a digital camera with which I've taken
> 3000+ pictures in the past two months (!).  I'd like to give each of
> those pictures an r1 URI so I can record information about the
> pictures.  These are mostly pictures of my kids, so I really need x1:
> I want the pictures and metadata to co-exist happily in 50 years.
> Also I want r2 URIs, so people can see the pictures!

If enough people want to cooperate with you to keep the pictures
available, you can use something like freenet. Otherwise, I don't
see why the Internet/Web should meet this requirement. There's
no such thing as a free lunch.


>   And r3 and r4
> URIs so my photo-browser applications can search for pictures and tell
> me about them, along with displaying them.  If all four URIs could be
> the same, it would sure make it easier to keep this huge pile sorted
> out.
> 
>     -- sandro
> 
> [1] http://www.w3.org/TR/2003/WD-rdf-mt-20030905/#urisandlit
> [2] http://www.w3.org/2003/08/owl-systems/test-results-out
-- 
Dan Connolly, W3C http://www.w3.org/People/Connolly/

Received on Wednesday, 24 September 2003 09:29:08 UTC