Boyce and Dalterio (1996) pointed out that electronic publishing means different things to different people. With regard to the classic publication model author-publisher-library-reader, they identified six phases of electronic publishing: author preparation and submission of the electronic manuscript, peer review, copy editing and typesetting, database preparation, production and distribution, and archiving. To librarians, access to and distribution of electronic publications as well as archiving (which will be dealt with in section 3.5.) are of obvious importance.
While book-like electronic media like CD-ROMs and diskettes are
processed in libraries very similarly to print media, remotely stored
electronic publications bring upon us changes that effect all phases of
collection development. Selection criteria for purchasing material
probably remain more or less as before (Nisonger, 1996), but it
might be more difficult to obtain evaluative information (e.g., book
reviews) about electronic material. In January 1997, the first journal
devoted to
providing such evaluative information was launched
and may find a lot
of approval among acquisitions librarians.
Once the decision to purchase (or lease) electronic publications has been made, there are often technological implications in trying to access the documents. Special software might be necessary that needs to be properly installed on local PCs or made available through the network. The library needs to provide public terminals and printers to ensure that patrons visiting the library can access the material. Because of the extensive use of the World Wide Web in the scientific community, it is hoped that a standardized way of providing electronic documents (i.e., in SGML (Standard Generalized Markup Language) or HTML (Hypertext Markup Language) format) will evolve that will overcome the various and confusingly different software packages currently used by publishers.
Cataloging needs to include all the bibliographic details catalog records for paper-based documents contain, plus information about access procedures, electronic format, size of files, hardware and software requirements and other technical details. Most libraries that have automated their catalog use the MARC (MAchine Readable Cataloging) formats for bibliographic representation. MARC currently is being enhanced by a subset of fields that cater to additional descriptions of electronic publications, including the electronic address of the document. Library catalogs with a hyperlinked World Wide Web interface can then provide access to electronic publications from within the catalog just by clicking on the address.
Access control can be taken care of by the publisher through IP (Internet Protocol) addresses, or it can be left as the responsibility of the subscriber (i.e., the library) to inform authorized readers about the user ID and password. The latter solution is not preferred by librarians, as it is nearly impossible to distribute a password to the library's user community and at the same time avoid any possible misuse by persons not entitled to access.
Costs of electronic resources will be difficult to calculate from now on. Publishers have come up with a large variety of new pricing strategies (see Brown and Duda, 1996); ``pay per view'' and ``flat fees plus additional charges according to usage'' are just two examples. Pricing models are still in flux and changes to current solutions are inevitable. In addition, different libraries may have to pay different prices for journal subscriptions. Some publishers have started to offer special conditions and reduced prices to consortia of libraries. Participants need to pay only a portion of the total subscription costs, but gain access to all the publisher's journals. For large university library systems, this model might be advantageous, but for small individual libraries it can be difficult or impossible to participate in a consortium for administrative, organisational or political reasons.
Instead of owning the material, there is a clear shift towards providing access to electronic documents (electronic journals in particular) just for a given time. Libraries are leasing electronic material rather than purchasing it. The necessary contracts, so-called license agreements, raise another set of questions (see e.g., Soete, 1997): What exactly are the rights and responsibilities of the parties of the contract? Are there limitations regarding simultaneous accesses, downloading or printing of material? What are the technical requirements? Is the library entitled to access the user statistics in order to evaluate usage or does the publisher keep these data secret?
One of the most critical parts of license agreements concerns
copyright. Publishers from many countries currently allow
libraries to take photocopies of their print publications based on ``fair
use'', i.e., if it is
for personal, not commercial, use. This naturally includes
interlibrary loan, but ILL can be restricted or
even prohibited by license agreements. A typical license currently favors
the information provider over the information user (Okerson, 1996a). If
the restrictive clauses
are not eliminated from the contracts, they will disable
traditional and necessary cooperation among libraries.
A positive exception is the text of the
American Astronomical Society (AAS) contract. The AAS
has included a librarian in their Publications Board for several years
and the license agreement they have set up for their electronic journals
is exemplary in its fairness and
thoughtfulness. The problem of
negotiating contracts for electronic documents has led to extensive
discussion among librarians. Documents that
guide librarians through the jungle of licensing, for example the
``Liblicense'' resource are necessary and
very welcome. A newly established mailing
list devoted to the
topic of electronic content licensing, liblicense-l, received some 800 subscriptions
within only a few days of its inception which clearly shows that
discussion and advice
on this topic are sought by many librarians.
It is the nature of leased material that it remains physically with the provider after the contract has ceased. Librarians must be aware that they may be left without anything when a license agreement terminates. This situation is unlike the print environment where a library at least owns the journal volumes that were published while a subscription lasted. Some journal publishers deliver backups of electronic journals on CD-ROMs which stay in the library even after cancellation of the subscription. In addition to the problem of whether or not CD-ROMs will be a viable long-term storage medium, they are not capable of presenting one of the major advantages of electronic over print journals, i.e., they are not interlinked with other journals or databases. Backup on CD-ROM therefore is not an appropriate alternative. The only fair solution would be to allow libraries access to those volumes they subscribed to.