[OAI-implementers] (Possible) Errors with OAI repositories

[Tim Brody]

(apologies for cross-post)

Hi all,

I've started harvesting OAI-PMH version 1.x repositories (so that they are
accessible to OAI 2.0 harvesters):
http://celestial.eprints.org/cgi-bin/status

If you have a repository (version 1.0, 1.1, or 2.0) and you're interested in
having it mirrored/gatewayed please email me with your baseURL.

I hope repo. admins don't mind me posting problems I've found here -
otherwise it would involve a lot of individual correspondance for fairly
common problems!

The following repositories appear to have XML character errors that prevent
parsing (according to Perl 5.8.0 XML::Parser):
217.21.43.5.ELibBSU
aim25.ac.uk
archives.anlc.uaf.edu.anlc
bis-oai.bis.uni-oldenburg.de.UnivOldenburgBIS
http://dispute.library.uu.nl/cgi-bin/oai.pl
documentation.in2p3.fr.in2p3
DSpace at MIT
elib.suub.uni-bremen.de.SUUB
elib.ub.rug.nl.RUGNL
IMAGEBASE.LIB.VT.EDU
lib.umich.edu
http://wo.uio.no/as/WebObjects/theses.woa/wa/oai
http://techreports.larc.nasa.gov/ltrs/oai2.0/
oai.library.uiuc.edu
www.biomedcentral.com.bmc
preprint.chemweb.com.CPS
www.numismatics.org:80.ans

The following repositories are not supporting yyyy-mm-dd format requests:
byu.edu:Ancestry
http://www.ifeanet.org/oai/oai2.php

The following repositories appear to be broken (error on OAI requests):
cds.cern.ch
nsdl.org

Misc:
etdindividuals.dlib.vt.edu.ETDIndividuals appears not to accept a resumption
token it returned
perseus.tufts.edu is responding with HTTP 400 errors for OAI-level errors
(should be HTTP 200 for all OAI responses)
www.davidrumsey.com:8080.Rumsey Collection appears to have broken
flow-control (returns the same token more than once)
xtcat.oclc.org seems to be very slow, may be a problem my end (big
responses?)
rib.cs.utk.edu Identify baseURL is set to localhost (causing much confusion
by my harvester :-)

All the best,
Tim