[OAI-general] Relative URI as identifier of OAI-PMH

[Young,Jeff (OR)]

Eiji,

What URI would this be relative to? Keep in mind that URI aliases would be bad for the intended purpose of unambiguous identification of the item across repositories.

As an aside, a Linked Data perspective puts a whole new spin on OAI-PMH. For example, why does OAI-PMH use Uniform *Resource* Identifiers to identify things they don't consider to be a "resource" (viz "items")?

"Note that the identifier described here is not that of a resource."

If we on the OAI-PMH technical committee could have foreseen the TAG httpRange-14 and genericResources-53 decisions, we might have tried harder to rationalize "items" as "resources". The result would have been more RESTful and your question would have been moot.

Jeff

-----Original Message-----
From: oai-general-bounces at openarchives.org [mailto:oai-general-bounces at openarchives.org] On Behalf Of Eizi TOYODA
Sent: Thursday, May 06, 2010 1:43 AM
To: oai-general at openarchives.org
Subject: [OAI-general] Relative URI as identifier of OAI-PMH

Hi,

Some people use just a filename (such as md0001.xml) for identifier of
OAI-PMH.  Would that be compliant to the protocol?  I really don't
like that, but the specification seems to be a little unclear to
persuade people to change practices.

The protocol (http://www.openarchives.org/OAI/openarchivesprotocol.html#UniqueIdentifier)
says:
"The format of the unique identifier must correspond to that of the
URI (Uniform Resource Identifier) syntax."
The generic syntax of URI includes relative URI, hence "md0001.xml"
might be a kind of valid URI.  But the overall context suggests
presence of URI scheme, so I wonder the community may not have been
aware of that.

Any thoughts of the community are greatly appreciated.

Best Regards,
--
Eiji (aka Eizi) TOYODA (Mr./PhD)
Japan Meteorological Agency
toyoda at npd.kishou.go.jp / toyoda at gfd-dennou.org

_______________________________________________
OAI-general mailing list
List information, archives, preferences and to unsubscribe:
http://www.openarchives.org/mailman/listinfo/oai-general