Re: In defence of 404 ...

Karl
> Maybe Bernard Vatant wants to express that geography nomenclature is 
> changing but not necessary references in published materials or even 
> in social context.
This is one aspect of the issue, yes.
>
> For example two cities being grouped in one and then one of them is 
> losing its name and its legal existence, but for history papers still 
> exist, etc. 
That's what I've tried to explain, but apparently with mitigated success.
> Geography is social, very social
Sure. Even the so-called "Physical Geography". See what Pat Hayes wrote 
recently about Mount Everest. (Sorry to quote Pat again.)
> there has been a very interesting paper about that
> ....
> http://etudesrurales.revues.org/document716.html
> Mon, 07 Aug 2006 13:19:46 GMT
>
> But Maybe Bernard is talking about something else
No, maybe I simply tried to make this an illustration of a more general 
issue. When you mint a URI (or for that matter, any kind of name or 
identifier) and publish it as a way to refer to an entity, you think 
about the state of this entity at the time of publication. But 
references generally last longer than their referent, and/or the 
referent properties can drift along with time. RDF specifications, and 
various recommendations and best practices, as they stand today, make no 
provision for taking this reality into account. FYI, in Mondeca ITM, 
since the very beginning of the application, we have the built-in 
ability to mark every entity, attribute and relation with a "beginDate" 
and "endDate" metadata allowing to filter the information valid in a 
given time span. I've always been frustrated not to have the ability to 
declare the same thing in RDF, using similar attributes directly on a 
RDF description, like :

<rdf:Description rdf:about="foo" rdf:endDate="1982-05-12">
    <!-- some elements of description of foo valid until 1982-05-12 -->
</rdf:Description>
...
<rdf:Description rdf:about="foo" rdf:beginDate="1982-05-13">
    <!-- some elements of description of foo valid since 1982-06-12 -->
</rdf:Description>

Bernard

Received on Monday, 7 August 2006 15:41:22 UTC