Eric Hellman wrote:
http://catalog.library.jhu.edu/bib/NUM identifies a catalog record- I
mean what else would you use to id the catalog record. unless you've
implemented the http-range 303 redirect recommendation in your catalog
(http://www.w3.org/TR/cooluris/), it shouldn't be construed as
identifying the thing it describes, except as a private id, and you
should use another field for that.
Of course. But how is a link resolver supposed to know that, when all it
has is rft_id=http://catalog.library.jhu.edu/bib/NUM ??
I suggest that this is a kind of ambiguity in OpenURL, that many of us
are using rft_id to, in some contexts, simply provide an unambiguous
identifier, and in other cases, provide an end-user access URL (which
may not be a good unambiguous identifier at all!). With no way for the
link resolver to tell which was intended.
So I don't think it's a good idea to do this. I think the community
should choose one, and based on the language of the OpenURL spec, rft_id
is meant to be an unambiguous identifier, not an end-user access URL.
So ideally another way would be provided to send something intended as
an end-user access URL in an OpenURL.
But OpenURL is pretty much a dead spec that is never going to be
developed further in any practical way. So, really, I recommend avoiding
OpenURL for some non-library standard web standards whenever you can.
But sometimes you can't, and OpenURL really is the best tool for the
job. I use it all the time. And it constantly frustrates me with it's
lack of flexibility and clarity, leading to people using it in ambiguous
ways.
Jonathan
