With the journals providing only the editing and peer review portions of their original functions, the issue of presenting and archiving the data needs to be addressed. Will there be one central archive, i.e. a ‘megacenter’ for the whole body of scientific knowledge akin to the Pubmed abstract archive, or will there be a system of federated archival libraries, e.g. the Biomed Archives Consortium66, Project Muse67, Highwire Press47, 68 or CrossRef69? Will it be privately (as is the case now with journals) or publicly controlled? Should the archive include only peer reviewed information, or gray literature as well?
One commonly used example of a central archive that has done exceptionally well is the physics preprint archive. In 1991 Paul Ginsparg launched this groundbreaking archive of physics preprints, http://arXiv.org (Formally operating out the Department of Energy's Los Alamos National Laboratory now working out of Cornell University). The archive, which receives tens of thousands of papers annually functions to rapidly and efficiently distribute articles as soon as they come out, even before they are published70.
While the international nature of scientific research would seem to make the concept of a centralized database politically unlikely 71. Still central archives have their proponents. Matt Cockerill of Biomed Central claims that it is imperative that data be stored within a central location for there to be efficient searches of the data. Additionally, a central repository can provide for a simple and interoperability friendly interface; fears of lost data can be limited if there are multiple mirror sites72. The costs of maintaining any long term digital archive favor a centralized archive over some balkanized system of small independent and non-interoperatable systems.
CrossRef, which aims to not only include journals but gray information as well such as, books, reference works, and databases, claims that they can achieve the same degree of interoperability, through the use of consensus standards, that a centralized archive can achieve, yet at the same time avoid many of the limitations inherent in a central system69.
SPARC (Scholarly Publishing and Academic Resources Coalition), is another example of a decentralized group. It is composed of universities that publish and archive an aggregate of leading research journals at prices that are ‘sensitive to the interests’ of publishers and subscribers accessible journals73, 74.
A digital archive in whatever final form it takes will have many advantages over the present day paper archives in libraries around the globe. For example, in contrast to present day libraries that cannot curate their physical stacks to remove wrong, misleading or outdated information, the dynamic nature of an online archive allows for the sequestering and possible removal of bad data. Moreover, similar to present online databases, the archive will be organic, growing and evolving based on the present and future needs of the research community.
The role of present day libraries will change from being physical repositories of information to being a ‘gateway of information’ providing advanced search systems and an ‘expertise center’ in terms of knowing how to access the different levels of the chain of information in the archives75.