[OAI-implementers] Up to 383 OA Archives now, 159 of them GNU Eprints

Stevan Harnad harnad at ecs.soton.ac.uk
Sun Feb 20 13:14:29 EST 2005


Dear Tim, Kat and OAI-implementers,

Several things.

First, I have done by hand the first imported set of OAI Base URLs
that Tim imported for me for
http://archives.eprints.org/eprints.php


It was worth the time and effort as not only did the
archive number rise to 363 and the Eprints subset to 159, but I
now have a good idea of the kinds of OAI-compliant archives there
are, and many of them are not relevant to OA: They are sites that have
merely made themselves OAI-compliant to increase their visibility.

That includes commercial product/service sites and University
Library catalgues and home pages -- both completely irrelevant
to OA and its goals (and introducing noise into OA-space:
Must remember: OAI is much, much wider than OA!).

Second, I would like to import more. What was that first list?
The UIUC registry (all of it?). 
http://gita.grainger.uiuc.edu/registry/

What else can we now import.

One obvious one is all of OAIster: I will go over it all by hand
and remove all the archives that are not full-text OA content.
(Kat, could we have a file with all the OAIster archives
URLs and OAI BaseURLs? I can take it from there.)

Third, Tim, I'm not sure why the totals are not tallying in
the Registry: We have 159 institutional Archives now, but
your analysis only covers 90 of them: Is that because the rest
are not in celestial? But than please let us do something about
that, to get them in!

Also, of the 159 GNU Eprints Archives, only 120 are picked up
by your analysis: celestial again? Important to maximise 
celestial's coverage!

Last thing: There are 4 remaining archives in the buffer:

edoc.mpg.de (I have that already, and it says so, but I can't
remove it from the buffer)

nsdlib.nsdl.cornell.edu -- that might be an important one, but I
couldn't get a connection to verify it

www.auborn.edu -- that is one of those annoying ones that pop out
of the buffer automatically: I need a "safe" viewer even for
the buffer stage, as with such archives I cannot make them "dead"
as it stand!

www.numdam.org is, like mpg, already in the registry, and it
tells me so, but it won't let me remove it from the buffer!)

Please import more OAI baseURL lists as soon as you can, 
especially OAIster

(And any autjomatic checking you can get it to do -- to see
is we have it already and de-dupe, would sure help minimise
my time and keystrokes!)

Cheers, Stevan




More information about the OAI-implementers mailing list