On Fri, 2004-10-22 at 06:30, Ray Denenberg, Library of Congress wrote:
> Yes, of course, sorry if that wasn't clear -- it's how we do it in SRW -
> there is a sub-authority component; the info registry registered 'srw' with
> LC as the registration authority for that namespace, and LC has registered a
> number of organizations, allocated them supspaces, and then these
> organizations register objects within their subspaces; see
> http://www.loc.gov/z3950/agency/zing/srw/infoURI.html (and Z39.50 has been
> doing an analogous process, with ISO object identifiers, for nearly 15
I looked at that, but I still don't see the actual "list" of elements,
just info on the structure of the identifiers. And I'm still confused
about how "organizations register objects" -- is there a registry for
the individual objects?
> Well in the srw case, you wouldn't want the info registry (OCLC, NISO,
> whoever) to have to respond to requests from University of Liverpool,
> Oxford, Koninklijke Bibliotheek, PICA, etc, for an SRW subspace (not an info
> namespace, an srw subspace), would we? How is this different?
> > To do otherwise would
> > require another level in the registry, and I think we're getting almost
> > into structural changes here.
> Don't follow.
Basically, it's that I see a difference between
Both make sense to me, but they are different implementations of "info:"
in my mind. In info:lccn, the namespace represents an identifier, and
there is one authority (LoC). It's very direct, very simple. The same
for info:oclcnum, info:doi, etc. And in the case of those three there is
control over who can create the numbers, but anyone can use one of those
numbers to create the URI form. Then there are info:sici and info:sid
where there is control over the rules for how to create them, but anyone
can create the actual value. The structure of the info URI is the same,
but I need to think more on whether I think there's a significant
difference. (I need a good taxonomy of types of identifiers,
In info:srw the namespace is a subspace, as you say, that is then
further divided by "authorities". Perhaps this is a semantic difference,
because the Namespace Authority has a specific meaning in info (from the
Who is entitled to register under info? <<
A Namespace Authority is entitled to register a recognized namespace
after suitable review.
A Namespace Authority is the body that owns and manages a public
And the draft says:
info-URI = info-scheme ":" info-identifier [ "#" fragment ]
info-scheme = "info"
info-identifier = namespace "/" identifier
namespace = scheme
scheme = ALPHA *( ALPHA / DIGIT / "+" / "-" / "." )
Now I admit that my ability to read these "BNF" type statements is very
weak, so I could be reading this wrong. But it seems to me that the
namespace cannot contain a slash. So with a URI like:
"xv" is the namespace, and that corresponds to a Namespace Authority,
which appears to be LoC, since LoC has registered the namespace "xv". So
what you are calling "subauthorities" are not Namespace Authorities in
the info definition of that. They are really organizations that LoC
allows to create xv values (presumably as long as they follow certain
rules). In essence, info could care less about what happens after "xv"
since LoC is the authority in control of that. How the "xv" namespace is
used is entirely up to LoC to define, much the way that NISO defines how
a SICI is created.
So I think what we are actually discussing here (and sorry that it has
taken me so long to get to this point) is not info:xv but is a proposal
for the creation of identifiers for items in authoritative lists. Much
like the SICI, there may need to be an info URI namespace for these, but
the real meat is that there needs to be a standard way to create the
(NOTE: Everything after "xv" constitutes the identifier portion of the
from the Draft:
where "ddc" is the "namespace" component for a Dewey Decimal
Classification [DEWEY] namespace and "22/eng//004.678" is the
"identifier" component for an identifier of an information asset
within that namespace.
This means that I would still like for us to address the problem set,
which I sent out before as:
1) We have a large number of independent lists that have to be
2) Some of these lists were developed for MODS, some are MARC lists, and
there are folks who probably want to create their own lists.
3) We want to make it easy to propagate these lists to users and to
4) We want to make it easy for humans to understand the lists and their
values, since they have to select the proper values from them when
5) We want to make it easy for programs to validate the values in
6) ?? add more here
None of this, nor even the identifier that we will create, has anything
to do with the "info" URI space. info can handle any identifier, and is
"blind" beyond the namespace level.
> "info:lc_xv" and "info:oclc_xv" versus "info:xv/lc" and "info:xv/oclc"
> (though many prefer "1" and "2" because the owners of the subspace change,
> but that's a different issue) functionally give you the same capability,
but there is a difference in meaning, because with "info:xv/lc", "lc" is
part of the identifier, not part of the namespace. That might not be
significant for the function you wish to serve, but there is a
> the difference is in the first case, a huge burden is placed on the info
> registry who probably doesn't want it; in the second case the burden is on
> an organization who has agreed to take it on.
So you are saying that the registry agency might prefer to have fewer
namespaces registered. I guess that will be a question for the permanent
agency when one is set up. But I think that since we don't know one way
or the other we should think about what is best for our community, and
hope that the agency will be able to handle that.
> so LC has (hypothetically) allocated a subspace to OCLC who registers
> "cartoon" within the subspace.
That still doesn't tell me where the individual values are registered.
They are not registered in the info registry. Do you really mean
"registered" or do you mean that OCLC has an authoritative list that has
"cartoon" as one value in it? And if so, how does one make the
connection between the URI and actual list?
Digital Library Specialist
Ph: 510-540-7596 Fax: 510-848-3913