URIs and OpenURLs (was:upcoming events for URI CG schedule?)

I'm listed as representing the library community to the URI Coordinating
Group. A while ago Chairman Norm was asking if there were any issues, etc:

Norman Walsh wrote:

> | Ray, any events in the library community we should know
> | about?

I do think I have something substantive to raise, the OpenURL effort.  Note
that Mark Needleman is listed as representing the OpenURL community to the
URI CG,  I have discussed this with him, and he suggested that it would be
better if I raised this issue;  my views on this do not necessarily reflect
his.  Anyway, here goes.

The OpenURL standard (in development) is of profound importance to the
library and publishing (and related) communities. There are a number of
references, but start at:
http://library.caltech.edu/openurl/
and there is a bibliography at:
http://library.caltech.edu/openurl/Bibliography.htm

There has been, far as I can tell, a near-total disconnect between the
W3C/IETF and the OpenURL community. I don't understand the reason for this
disconnect and I'm not attempting to address the big picture and identify or
fix the underlying societal problem (if there is one), but I want to raise
conciousness on one particular potentially-serious issue.

There is a proposed identifier framework, at:

http://library.caltech.edu/openurl/PubComDocs/Announce/20030415%20Announce-Naming.htm

While a quick read this document might raise a number of concerns,  the most
serious, I think, is the suggestion that URI is one of a number of so-called
"naming environments" -- specifically, three: URI, ORI, and XRI (although it
isn't clear if this proposed framework is intended to be extensible to
additional "environments") and that every identifier is to be prefixed by its
environment type.  Thus (and most significantly) it is presumed that URIs
will be prefixed by 'uri' (as in "uri.http://xxxxxx").

Now I've told everyone I can (who will listen) that nobody's going to start
prefixing their URIs in this fashion. And this discussion is strongly
reminicient of the debate over URLs and URNs a number of years ago, when
there was an analogous  suggestion based on the premise that URL and URN were
two of possibly several naming environments and that URLs and URNs  should be
prefixed by 'url:' and 'urn:' respectively.  As we all know that suggestion
wasn't adopted. The proposal in question risks reviving that earlier debate,
which (in my opinion) would be a huge waste of time and energy. But it's
instructive to recall how that earlier debate was eventually resolved: 'urn'
was adopted as a URI scheme.

I really don't think the OpenURL folks honestly believe that their scheme
will be adopted; I think they've put it up out of some frustration and
cynicism. I don't completely understand why they think these things that they
want to identify cannot be accomodated within the URI framework either with
one or two new URI schemes or new URN namespace identifiers.  However I
suspect it is because they have concluded that it is nearly impossible, or at
least very difficult, to get new schemes or namspaces approved.

I would like to hear what others on this list think about this. I would
appreciate it if some of you might be willing to study this and provide
guidance.

--Ray