Re: How do you flag a resource which is not available anymore?

On 6/2/14 12:47 PM, Bernard Vatant wrote:
> Karen, Kingsley
>
> Sorry, rant ahead :)
>
> HTTP URIs are both identifiers and locators and there are thousands of 
> pages expanding at length on what they mean or not, etc. I don't want 
> to go again in this permathread of URI meaning. Just want a name to 
> flag clearly a situation which will be more and more frequent, and 
> which semantic web actors just plainly ignore or even contribute to 
> create. See below.
>
> Please look at just the following example, and tell me what is your 
> concrete proposal, beyond all theory.
>
> http://lov.okfn.org/dataset/lov/details/vocabulary_geof.html
>
> A backup version of the vocabulary is available at 
> http://lov.okfn.org/dataset/lov/agg/archives/dir_geof/file_geof_2004-02-20.n3
>
> Now look at what the vocabulary 
> http://www.mindswap.org/2003/owl/geo/geoFeatures20040307.owl is 
> currently dereferencing to.
> It's 100 times worse than having a 404. It's nothing to do with 
> temporary failure, mindswap.org <http://mindswap.org> which used to 
> host a wealth of precious resources for the Semantic Web (see 
> http://www.cs.rpi.edu/~hendler/presentations/Harvard-IIC-05.pdf 
> <http://www.cs.rpi.edu/%7Ehendler/presentations/Harvard-IIC-05.pdf>) 
> is currently a domain to sale with a garbage placeholder for any URI 
> http://www.mindswap.org/whatever-goes-here. And the devil knows who is 
> going to buy it tomorrrow and what the content will be.
>
> OK the future is unknown, but in that case I can bet anything you want 
> that this domain will never ever come back to its original owner and 
> the URIs as geof restored in line, because the original domain owner 
> is linked to a project which is over for quite a while.
>
> I would be curious to have the take of Jim Hendler on this, since he 
> was the leader of mindswap project. Any chance to catch the domain back?
>
> See also aktors.org <http://aktors.org>, which is for sale, and down 
> the semantic drain goes AKT ontology, with all the data using it after 
> 10 years online
> http://lov.okfn.org/dataset/lov/details/vocabulary_akt.html
>
> What I want is a quick-and-dirty way to show this to the face of the 
> world.
>
> Best regards
>
> Bernard

To cut a long story short, I am not against a relation that provides 
additional clarity about an entity denoted by an HTTP URI. To good news 
is that we already have wdrs:describedby relation [1] for this very 
purpose. Even better, It also has an inverse in the form of 
xhv:describes [2].

Example:

## Turtle Start ##
<>
a foaf:Document;
rdfs:label "Loosely coupled Ontology denotation example" ;
rdfs:comment """Example of how to denote an Ontology for long term 
durability in regards to HTTP based data access. """ ;
foaf:topic <#someOntologyURI> .

<#someOntologyURI>
a owl:Ontology ;
rdfs:comment """An Ontology description where  denotation identifier is 
explicitly associated with the location of its descriptor (or 
description document)  i.e., actual
                           RDF document comprised of RDF statements in a 
specific notation e.g., TURTLE """ ;
wdrs:describedby <OntologyOldLocation.ttl>, <OntologyNewLocation.ttl> .

## Turtle End ##

Links:

1. http://www.w3.org/TR/powder-dr/#assoc-linking -- wdrs:describedby
2. http://tools.ietf.org/html/rfc6892 -- xhv:describes .

-- 

Regards,

Kingsley Idehen 
Founder & CEO
OpenLink Software
Company Web: http://www.openlinksw.com
Personal Weblog: http://www.openlinksw.com/blog/~kidehen
Twitter Profile: https://twitter.com/kidehen
Google+ Profile: https://plus.google.com/+KingsleyIdehen/about
LinkedIn Profile: http://www.linkedin.com/in/kidehen