Sorry to say it, Peter, but what you're trying to do can't be done -
indicating the type of identifier using DC simple.  And trying to guess the
type heuristically is a bad idea in my opinion. And establishing a
convention for use of dc identifier in SRU, just as bad. I would not
advocate either approach.  Sorry, wish I had a suggestion.  As to the
suggestion that we should establish a DC qualified schema, I wouldn't want
to do that unless someone says they need it.   If your "partner" decides
they want to use DC qualified we can work that up fairly quickly I would
think. (But then, why not just use MODS?)

--Ray



-----Original Message-----
From: Z39.50 Next-Generation Initiative [mailto:[log in to unmask]] On Behalf Of
Peter Noerr
Sent: Wednesday, October 12, 2005 7:00 PM
To: [log in to unmask]
Subject: DC:identifier semantics

I'm going to believe that that is a good thing, Sebastian, though not so
much of the old age please! It did give me a chance to add a subject line
which I forgot the first time. (No 'senior moment' comments please.)

> -----Original Message-----
> From: Z39.50 Next-Generation Initiative [mailto:[log in to unmask]]On Behalf 
> Of Sebastian Hammer
> Sent: Wednesday, October 12, 2005 5:48 PM
> To: [log in to unmask]
> Subject:
>
>
> I do believe Peter is getting all warm and fuzzy-like in his old age. 
> Good for you, Peter!
>
> --Sebastian
>
> At 05:19 PM 10/12/2005 -0600, Peter Noerr wrote:
> >A probably really dumb question:
> >
> >A partner of ours wants to use a URL as a "dc:identifier" for
> their records.
> >No problem so far this is one of the recommended values. They
> want it to be
> >used as the link to read the record. Also no problem .
> >
> >However as a metasearch engine we need to be able to identify
> that it is a
> >URL and a link within a set of records from a number of different 
> >Sources (for the sake of being awkward - like real life - assume all 
> >are
> queried via
> >SRU and are returning records in simple DC). They don't want to go 
> >beyond using simple DC, so qualification is not an option. Using a
> "urn:url" type
> >namespace identifier would work. If everybody used it, which
> they won't, we
> >would be home free. [What is the 'namespace' identifier for a
> URL when used
> >within the DC set in SRU anyway? ISSNs ("urn:issn:") and DOIs
> ("info:doi/")
> >are all listed, but URLs are not]
> >
> >Are we limited to detecting "http://" and deducing that it is
> not an ISBN?
> >Of course some will start "www." and others will be just
> "anywhere.com/..."
> >So this is a bit of a nightmare.
> >
> >I don't want to have to produce hard coded variants of our SRU
> Connector to
> >handle this sort of issue. I want to keep a single Connector which is 
> >configured for different Sources - if it has to be. The less
> variation there
> >is the less configuration there is and the more likely things
> are to work.
> >
> >So the (main) dumb question is: How do they put this link URL in
> a returned
> >record in simple DC in a way that it can be recognized by dumb 
> >computer programs? Or are we stuck with configuration?
> >
> >
> >Peter
> >
> >Dr Peter Noerr
> >Chief Technical Officer
> >Museglobal, Inc.
> >
> >tel: +1 801 208 1880
> >fax: +1 801 208 1889
> >cell:+1 801 910 4912
> >
> >[log in to unmask]
> >www.museglobal.com
>
> --
> Sebastian Hammer, Index Data, www.indexdata.com Direct phone: (603) 
> 209-6853 Fax: (603) 357-1813
>