In order to guarantee continued access to information, publications have been archived throughout time. Traditionally, preservation has been a task of libraries, which they have performed reliably through the centuries. Up to now, even small specialized libraries were able to provide a highly valuable repository of contemporary and historical publications. Their holdings could be unique within their geographical region.
We have seen many media tested for and applied to archiving, for example paper, microfilm, microfiche, CD-ROM, and magnetic tape. Of all these, paper still is the preferred storage medium for many librarians, since it is the only format that can be used without any special viewing devices and therefore is independent of current and future technology. In addition, no other storage medium up to now has survived for as long as paper.
Electronic publications turn this ordered situation upside-down. They unfold their most valuable features when they are interlinked with other documents, archives, and databases. Thus, if we regard electronic publications not as self-contained, off-line documents, but as parts of a large system, we must conclude that they need an archiving system that is based on a global concept and goes beyond our current model.
A new physical storage medium (e.g., magnetic tape, networked on computers) as well as an appropriate data format (e.g., SGML) have to be selected (Wusteman, 1996). It is obvious that current technology soon will be replaced by newer versions or entirely new systems, therefore both the physical storage medium and the data format must be sufficiently flexible to be transferred to the next technological generation. This transfer is being referred to as ``technology refreshing'' or, as an even broader concept, ``technology migration'' (Garrett and Waters, 1996).
The rapid pace at which technology becomes obsolete makes archiving an extremely expensive undertaking. Libraries' budgets are highly unlikely to rise with the necessary speed (in fact, they are unlikely to rise at all). Librarians may find themselves being forced to abandon archiving, one of their traditional tasks. New volunteers appear on the scene, some eager to take over. In addition to non-profit organizations, commercial institutions and subscription agencies participate in the debate. Many librarians are skeptical about whether the interest of commercial organizations in archiving will go beyond immediate economical consideration. We also wonder whether we will have to pay twice for journals in the end - once for the license, and again for access to back issues.
In contrast to commerce-oriented solutions, the Task Force on Archiving of Digital Information, established by the Commission on Preservation and Access and the Research Libraries Group, suggests in its final report that a national system of digital archives should be developed (Garrett and Waters, 1996). Digital archives in their sense are ``held together in a national archival system through the operation of two essential mechanisms'': the repositories must meet or exceed criteria set by a program for archival certification, and a ``critical fail-safe mechanism'' must be in place that can carry out rescue functions when needed. This report focuses on the United States and although such a well-organized, nation-wide archival system will be difficult, maybe even impossible, to achieve for many countries, the Task Force envisions a possible expansion on an international level.
International standards for archiving electronic publications in conjunction with a stable and mature network environment would be the ideal solution, but it will take some time before the necessary infrastructure will be in place. In the meantime, we must find other solutions (see e.g., Neavill and Sheblé, 1995). We have reached a critical point in time, because technology has become so advanced that publishers are tempted to abandon the traditional archiving medium before actually being able to come up with another long-term solution that is appropriate for the digital age. If the mere availability of technology leads to short-sighted solutions, it will have expensive, if not fatal implications for future access to scientific literature. In this regard, the opportunities given to us by technology clearly are a danger. In order to ensure uninterrupted access to publications, archiving must continue on paper until the question of infinite preservation is fully clarified.