Differences among electronic journals are remarkable. Publishers may label a publication "electronic" even though it is just an online clone of a paper edition; these journals merely reproduce the printed page in electronic format, neglecting features of electronic publications that really make them superior to paper, for instance extensive links within documents, in particular from references and figures back to the text, as well as to other documents and data resources. Readers seldom demand that publishers enhance their journals as many seem to be content to have at least something accessible from their desktops, provided the access speed is acceptable. From communications with astronomers we know that the ability to print articles still is highly important. The urgent need for quality electronic publications with significantly increased functionality and, even more important, archives in robust formats like SGML (Standard Generalized Markup Language) and XML (Extensible Markup Language) that allow translation into any new standard, remains to be fully recognized by the astronomical community (Boyce, 1997). If a resource is popular for whatever reasons, few users seem to worry about possible improvements or future availability. Librarians are mediators between authors/readers and publishers, and they can play a vital rôle in alerting both parties to values and dangers intrinsic in high or poor quality electronic journals.
The electronic environment changes immensely how publications can be used today and in the future. The traditional copyright is being replaced by license agreements which have to be negotiated and, finally, signed for each publisher, sometimes even for each individual journal. Librarians have to obtain considerable legal knowledge in order to protect as many existing user rights as possible against the restrictive, market-oriented approach of some publishers. Mailing lists and web sites like LIBLICENSE (http://www.library.yale.edu/~llicense/index.shtml) and EBLIDA (European Bureau of Library, Information and Documentation Associations, http://www.eblida.org/), dealing with licensing issues, are to be praised for the support and advice they provide. To the users of electronic journals, the complex licensing procedures are seldom apparent; the only trace often is a header on the e-journal's page or, in only few cases, a screen that appears once per session, informing readers that "this subscription is registered to the XXX library". Libraries appreciate these notes; often they suggested including them as they help to alert users to the ultimate source (at least the financial one) of electronic publications.
Much has been said about the science journal crises. Subscription costs are continuously rising, partly due to an ever increasing number of pages per volume, partly because of the publishers' attempts to meet their shareholders' interests, and, most importantly, electronic versions of existing journals are not available for free, but can increase the subscription price by 20%. The largest portion of our budgets are for journal subscriptions, and little flexibility is left when it comes to book purchases. To make most efficient use of any remaining money, a carefully considered collection development policy is more important than ever, otherwise the increasing demands from users cannot be satisfied with decreasing budgets.
The LANL astro-ph e-Print server currently is the most frequently used astrophysics preprint database; an increasing number of articles cite astro-ph reference numbers, and to many scientists astro-ph submissions are as good as published ones. However, at the time of their submission to astro-ph, they aren't yet published, and before they are, title, authors, and contents may be modified. Various versions of e-prints can co-exist, and unless the authors care to update submissions by adding the final bibliographic reference, papers can remain marked as preprints forever even though they may have been published in the meantime. The unique reference numbers assigned by e-print servers seem to lend themselves to automating the updating process, but at present these tools are not yet in place. The only reliable databases that allow follow-through from preprints to published articles still are the Space Telescope Science Institute STEPsheet (http://NTweb.stsci.edu/STEPsheet/)and the National Radio Astronomy Observatory RAPsheet (http://annie.cv.nrao.edu/rapsqbe.htm) which are maintained by the STScI and NRAO librarians respectively.
By now the ADS Abstract Service has become the de facto standard reference tool in astronomy. Many astronomers start their literature searches at ADS where they retrieve articles published in all major journals, often with the option to "click through" to electronic versions of papers and to cited as well as citing articles. ADS usually gets permission from publishers to link to full texts of e-journals from their database. However, for most conference proceedings including IAU symposia and colloquia and volumes of the ASP conference series, full texts are not yet available in electronic format. In order to nevertheless make these references available through the Abstract Service, librarians assist ADS by copies of contents tables and title pages upon receipt of the published volumes.
Many library activities don't immediately hit the eye of scientists and, ironically, users seem to look at libraries most if they are not satisfied with our service. This dilemma is known to everybody who ever tried to run a household; hardly any family member cares to commend a tidy flat, but don't you dare to be sloppy with daily chores - you will get complaints instantaneously. Librarians can react to this "invisibility" in two ways: we can remain quiet and hope that someone, finally, will recognize the value of our existence. Or we can try to increase our presence in the daily life of scientists by actively emphasizing our rôle. There are many ways to advertise library services.