[sc34wg3] Fwd: [httpRange-14] Resolved

Lars Marius Garshol sc34wg3@isotopicmaps.org
Tue, 12 Jul 2005 14:09:25 +0200


* Steve Pepper
| 
| Really? The whole thing seems to me to be totally at odds with
| the subject identifier approach:
| 
| | |   a) If an "http" resource responds to a GET request with a
| | |      2xx response, then the resource identified by that URI
| | |      is an information resource;
| 
| A subject identifier that follows the OASIS recommendations for
| published subjects *will* give a 2xx response because it will be an
| http URL that resolves to a subject indicator. That subject
| indicator will usually indicate a subject that is *not* an
| information resource.

I think there are two cases here:

  1. PSIs

  2. URIs for resources which are not information resources, defined
     by others


In the case of PSIs the URI is the URI of an information resource. To
take an example

  http://psi.oasis-open.org/iso/3166/#124

is the URI of an information resource that *describes* a country. So
in indirect identification schemes (like [subject identifier] in topic
maps), it can be used to identify a country instead of the information
resource, and if used in [subject locators] it actually does identify
the information resource. 

It always remains the URI of an information resource, however, and so
I don't see any problem here.


In the second case the URI, if resolved, produces an HTTP 303
response. From the point of view of topic maps this isn't so
interesting, because if it appears in a [subject identifier] property,
what the URI resolves to is only for humans, anyway. It means it's a
less than great PSI, but everything will work nicely (especially if
the redirect goes to something human-readable).

If used in [subject locators] attempts to resolve it will give HTTP
303, which indicates one of two things:

  1. The real information resource resides somewhere else, for obscure
     reasons. (This is the original use of 303, according to RFC 2616.)

  2. There is no information resource, and the whole topic is pretty
     suspect.

This is really "the web as usual", and nothing to worry about.


I guess another way to describe it is that from the point of view of
topic maps the idea that URIs identify something on their own is not
valid. URIs can be resolved, but to know what a URI identifies one
must know in what way the URI is used.

-- 
Lars Marius Garshol, Ontopian         <URL: http://www.ontopia.net >
GSM: +47 98 21 55 50                  <URL: http://www.garshol.priv.no >