The AAS Executive Office recognized the possibilities of electronic dissemination three years ago and embarked upon a plan to utilize the emerging Internet technology and software in distributing information of interest to the astronomical community (October 1992 AAS Newsletter or URL http://aas.org/newsletter/nlinsert). In developing the AAS plan, we used several guiding principles:
The AAS has a dual role in proceeding with an electronic publishing effort. The first is to ensure quality through rigorous peer review to validate the science and copy-editing to ensure clarity of expression. In the sea of unrefereed manuscripts, conference proceedings and internal publications, we see the journals as a repository of high quality science. The second role is to ensure that the astronomical community continues to have inexpensive and easy access to the research results published in the journals.
Publishing has many facets, and all have to be addressed. The heart of an electronic publishing system is the database of manuscripts to which sufficient logical markup has been added so that different types of output can be obtained from the same database. For instance, from a manuscript database, we need to produce output which is suitable for display on a screen, a file which can drive the printer at a local workstation and a file suitable to drive the typesetting software to produce the traditional, bound paper copy of the journal. To understand the degree of planning needed, consider that electronic publishing requires the following operations:
Further, all of these steps have to be done in a demanding production environment. The ApJ consists of more than 20,000 pages per year. Put another way, the process must produce 100 new pages every working day. Changes to the AAS publishing operation have to be done with great care as we can not afford to disrupt the production process. Careful and thorough planning is essential.
The scholarly community has been developing and refining the print process for several centuries. It is unrealistic to expect to be able to replace that body of experience in one step. We consider the move to electronic publishing to be an experimental effort in which we will be using new technologies, seeing what works and building upon the feedback and suggestions from the user community. The best way to gain experience and uncover problems is to start with smaller scale projects than the full journal. Good examples or this are the electronic meeting abstract system and the CD-ROM for data, both of which have taught us a great deal.
Today's technology and software tools are changing with a rapidity which we have never experienced before. Large (and expensive) systems based on special, proprietary software are doomed to fail. Even if they work, new technology will render them obsolete within months. The scale of the AAS publishing process requires a large effort, and fortunately there are ways to incorporate flexibility into such a system. Achieving flexibility requires adherence to open standards for uniform exchange of data. By conforming to accepted standards for data exchange we will make it possible for the community to have full access no matter what tools they will use. Understanding what is meant by standard is important. Standards for data access and interchange are different from a standard program. The first expands possibilities - the latter limits development. The AAS adopted standards which allow the data, i.e. the manuscript, to be accessed, searched and retrieved by any software tool which adheres to the standards. In this way, the data will be usable, but the access and display tools can change and grow without rendering the data unusable.
The AAS has adopted the following standards:
These standards are not static. Within the next year, new additions to Word Perfect, Microsoft Word and other commercial word processors are expected which will be able to produce output in SGML-tagged format. There will also be one or more WWW software products which will work with Mosaic and display SGML tagged documents, including advanced math. Both types of product will greatly simplify the task of producing and displaying papers with the complicated built-up math which is typical of astronomy manuscripts. As such new products become available the AAS will incorporate them into the electronic publishing process.
There are a number of excellent free software tools for browsing, searching and retrieving information from the WWW. The AAS is using public domain packages to serve information to the net and is structuring the information to be used effectively with free software such as Mosaic. We have specifically avoided solutions which depend upon proprietary software which could lock the astronomical community into undue dependence upon a single software manufacturer.
We now have new technologies, new tools, and new possibilities. The most important advantage of electronic publication will come from using the new capabilities in ways which we can only begin to imagine. Electronic information will be integrated into the user community in ways we can not yet envision. As producers and users of information, we have to think in new ways. We invite the AAS membership to help us develop new solutions to old problems, provide new ways of communicating and find new and more productive ways to enhance our quest to understand the universe.