SUMMARY: CERN has been one of the scholarly/scientific world's leaders in designing and implementing one of the first, and still the biggest and most successful institutional self-archiving policies: an institutional self-archiving mandate. As a result, CERN's Institutional Repository (IR) is well on the way to housing -- and thereby providing Open Access to -- 100% of CERN's own current published research article output (and soon its past output too). CERN has also designed CDS Invenio, a software package for creating Open Access Institutional Repositories. CERN is promoting its software (for "digital library management") but not promoting its exemplary self-archiving policy: Instead of first making sure that the rest of the research world follows CERN's own sterling example by self-archiving 100% of each institution's own respective research output, so as to make 100% of research worldwide OA, CERN is busy promoting the conversion of the world to Open Access publishing! Charity may begin at home, but it does not end there! CERN should be helping to convert the world to 100% OA before trying to reform publishing. If and when CERN does put its considerable weight and energy behind propagating its self-archiving policy to the rest of the world, then CERN and its CDS invenio software too will have earned for themselves, alongside the University of Southampton and its GNU EPrints software, the status of "World's Best Practice" infsofar as OA IRs are concerned.
On Fri, 28 Jul 2006, in response to my
posting -- to the effect that "EPrints is
free [and] the world's
first,
most widely used" [software for creating Open Access Institutional Repositories] --
Jean-Yves Le Meur commented in the
American Scientist Open Access Forum:
J-YLM: "This is indeed true in terms of number of installations [but] in terms of number of documents in repositories, CDS Invenio distributed by CERN [formerly CDSware] easily comes first... "
Jean-Yves is absolutely right that CERN's
CDS Invenio (formerly CDSware) comes
first among the top dozen digital archiving software installations in terms of average number of documents (28,327):
SOURCE: Registry of Open Access Repositories (ROAR)
Software #Archives Total#Items (Average#Items)
GNU EPrints: 210 194563 (1019)
DSpace: 162 288074 (2667
Bepress 51 96246 (2048)
ETD-db: 23 301394 (15863)
OPUS 21 6300 (350)
DiVA: 14 9453 (675)
CDS: 9 169963 (28327)
ARNO: 6 157754 (26292)
HAL: 4 39259 (9815)
DoKS: 3 1970 (657)
Fedora: 3 6946 (2315)
EDOC: 2 49687 (24844)
MyCoRe: 1 1984 (1984)
Others: 212 3220717 (24216)
I would go even further, and point out that CERN's own IR, with
78,774 items, is
11th biggest among all 721 archives registered in ROAR and
3rd biggest among the c. 450 of the archives that are institutional repositories IRs (rather than central repositories, which draw from the contents of many institutions), after #1
Wageningen University (110,269) and #2
Nagoya University (82,625).
Moreover, among IRs,
CERN's own IR, with its 78,774 items, mostly full-text, is indeed easily and by far the first (biggest and most successful). (Wageningen's admirable IR has 110,269 items but only about half of them are full-text documents, and the nature and percent full-text of Nagoya's imposing IR, with 82,625, still needs to be ascertained: a posting from Nagoya would be most welcome!).
I will go still further: CERN leads the world in Open Access IR "
Best Practice" Policy, being the institution with the most comprehensive, systematic and successful
institutional self-archiving mandate (and one of only
6 institutions worldwide that have a self-archiving mandate at all!).
The only problem is that CERN is not promoting the adoption of its superb institutional self-archiving policy along with its promotion of its CDS software!
CERN is instead too busy trying to
reform publishing (so as to hasten a transition to Open Access Publishing), having already successfully reformed its own researchers' self-archiving practices.
But "
Best Practice" insofar as OA (which is not the same thing as OA Publishing) is concerned may well
begin at home, but it must not '
end at home!
I would be more than happy to endorse CDS Invenio as OA IR Best Practice, alongside GNU EPrints -- if and when CERN promotes the adoption of its own institutional self-archiving
policy model along with its CDS software.
Otherwise CDS Invenio is, and will continue to be, just another of the softwares that (in its own words) "
covers all aspects of digital library management."
For OA IR "Best Practice" is not "to cover all aspects of digital library management": it is to focus specifically on the urgent OA priority: ensuring that 100% of institutional research output is systematically and successfully self-archived in the OA IR as soon as possible,
just as CERN's is. The outcome is already long overdue. CERN has already attained it. It is time for CERN to help the rest of the world to attain it too, by promoting its own institutional Best Practice along with its (excellent) software!
Meanwhile, that is precisely what GNU EPrints is doing, and has been doing all along: promoting not the practice of "digital library management" but the
practice of OA self-archiving.
Prior American Scientist Open Access Forum Topic Threads:
"CERN's historic role in OA" (Aug 2005)
"Eprints, Dspace, or Espace?" (Feb 2003)
Stevan Harnad
American Scientist Open Access Forum