today | current | recent | random ... categories | search ... who ... syndication

Tuesday, July 22 2003

It should be noted that we were almost called Borelia

(as in the North Wind to compliment the Southern one in Australia (ah, Empire...)) but for Queen Victoria's dislike of the name.

refers to


Excerpted : "The wonkish waters of RDF mailing lists"

Subject: [N3] equivalencies

From: Aaron Straup Cope 

Date: 22 Jul 2003 15:01:10 +0000

I'm wondering if you can answer a question for me and save me the

trouble and wading in to the wonkish waters of one or more RDF mailing


The question is premised on two assumptions :

1) The RDF that describes a thing is *not* public. That is I do not want

to share it and make it available to some other bot scraping the

network. If that makes a me a bad citizen, I'll live. 

All of which means I use URNs to describe things:

 @prefix uwh: \

   <urn:aaronstraupcope:knows:who:> .

2) At some point, I need to be able to resolve all that gibberish. I

need be able to tell the processor about something like this:

 @prefix awh: <user:pswd@>

Or simpler yet :

 @prefix awh: <file:/home/asc/knows/who/> .

Still with me? Here's the question. Does the spec DWIM (Do What I Mean)

when I say the following:

 uwh: = awh: .

That is, will a fully compliant processor be able to figure out that

when it comes time to merge a bunch of RDF documents will fetch stuff

from awh: namespace when it encounters things in the uwh: namespace?

If I feed what I've described to cwm I get the following:

 <rdf:Desription rdf:about="urn:aaronstraupcope:knows:who:">




    "user:pswd@" />


So it validates. But do I have to specify an equivalently for each

property (e.g. uwh:asc, uwh:bob) or does the spec just, well, DWIM?



Monday, July 21 2003 ←  → Wednesday, July 23 2003