[OAI-implementers] Notes on OAI 2.0 and SOAP
Wed, 06 Feb 2002 12:11:14 -0700
Thanks for the response on the OAI 2.0 schedule and process (contained
in messages from Mr. Van de Sompel and Mr. Lagoze). Before posting my
previous comment, I visited the OAI website and it appeared there that
the OAI 2.0 specification process was just beginning. I was not aware of
how much had already been discussed and decided. Nevertheless, I do have
a few additional comments to make.
1. OAI needs to get out of the web service infrastructure building
business and as Mr. Underwood states focus on metadata access. Perhaps
the resources that have been focused on custom protocol building could
be refocused to tackle the community building that Mr. Krichel refers
to. Misplacing resources on an obsolete protocol and the infrastructure
around it only exascerbates that problem.
2. The mantra of "simple to implement" seems to me to be something of a
red herring. OAI obviously realized early on that, for some institutions
if not most, no protocol would be simple enough to implement and that if
OAI wanted to reach those institutions they would have to provide
working implementations. Isn't that the reason that packages like
eprints, Kepler and the other OAI tools were produced? So the universe
of organizations that actually have to implement the protocol is
actually quite small. And I'm fairly sure that any of these institutions
have the technical ability to implement either protocol.
3. Finally, I strongly disagree with the restatement of the OAI mission
statement, limiting its goals to dissemination of information within
participating digital libraries. This is not the premise that is
promoted to repository providers. And it is certainly not a limitation
that I want for our repositories. I would add to Mr. Underwood's
statement that custom protocols are never simple, the statement that
custom protocols are never really "Open" either. They always create real
barriers to entry (e.g. Mr. Underwood's statement that it would not be
worth it for Inktomi to build a harvester for one-of-a-kind XML
protocol). As such, this is not a system being designed even for the
benefit of the limited group of repository providers (it certainly isn't
to our benefit or the benefit of potential searchers of our
repositories). And the net effect of this design is to replace one
resource oligopoly with another. Perhaps you ought to rename it the
SoOAI for "Sort of Open Archives Initiative" Or maybe NROAI for "Not
Really Open Archives Initiative". Just kidding.
The question is not whom on the technical, steering, or executive
committees or the participants of this listserv did or did not express
interest in whatever protocol and when. The right question is what
protocol potentially offers authors who contribute to these repositories
the widest possible resource discovery potential for the publications
they contribute. I don't think that, if they were asked, authors would
express much interest in any agenda other than that.
I know that there have been discussions in the past here about how to
generate the very interest from organizations like Inktomi that Mr.
Underwood seems to be offering. I don't understand now why you don't
jump at the opportunity when it is offered.
Health Sciences Center
University of New Mexico