[OAI-implementers] ABOUT & Schemas & metadataPrefix

Tim Brody tdb198@ecs.soton.ac.uk
Mon, 2 Apr 2001 17:28:57 +0100 (BST)


On Mon, 2 Apr 2001, herbert van de sompel wrote:

> > >From a harvesters point of view (I don't believe there are many of us :-), I
> > would prefer to have "oai_dc" because that tells me explicitly what data I
> > can expect to find, rather than having to remember what I requested (as far
> > as I can tell it is the one part that makes an isolated OAI response
> > stateful, a real pain if one is using caching or other systems).
> > 
> 
> The metadataPrefix is only signficiant within the realm of a certain
> repository.  The only exception is the metadataPrefix oai_dc, which --
> by convention -- refers to metadata expressed in unqualified Dublin Core
> in all repositories.  

> * what REALLY tells you which metadata you receive is the namespace:
> xmlns="http://purl.org/dc/elements/1.1/" .  that is a global identifier
> of the format.

Agreed, I guess when using OAI responses I need to delve into the XML
schema identification to be correct ...

> * in addition to that, you do not have to "keep" the format you asked
> for: all response are self-contained, meaning you can tell from the
> original protocol request -- which is the content of the requestURL
> element -- what you asked for.  this has been a deliberate choice, since
> we are indeed talking about robots harvesting metadata, and some
> software processing the harvested metadata at a later stage.

I have thought of this, but the protocol specifically discards this
information when using resumptionTokens (I, personally, don't like the
idea of exclusive variables vs the repository doing something intelligent
when asked something stupid - introduces state information that isn't
naive to implement on either side).

Of course, the namespace makes this immaterial.

Thanks for the quick response,
Tim.