The American Physical Society (APS) and the Standard Generalized Markup Language (SGML)

Archive of discussions on SGML issues

Direct access to the archive of discussions on SGML issues


Bob Kelly, writes:

The often stated objective of The American Physical Society (APS) is the advancement and diffusion of the knowledge of physics. In the last few years, a major goal of the Society has been to move its scholarly publications away from print and toward electronic production and dissemination. The current explosion in the use of information technology for early dissemination of scientific information dovetails with APS's interest in electronic publishing and consumption of scholarly journals and provides both physicists and the Society and other Physics Publishers and Librarians with an opportunity to join forces to support and improve scholarly communication in physics.

SGML is the center piece of the APS strategy to accept manuscripts electronically and to provide storage and delivery choices. With the advent of software, it is becoming feasible to make the SGML files of individual articles available for viewing.

It is believed that, if we as publishers can agree on some degree of SGML standardization, the authors and readers will benefit. A common SGML approach will facilitate ease of integration of papers from multiple publishers at the library or reader level. A common SGML approach will facilitate the development of many authoring tools and reading choices.

The term "Co-opatition" has been coined to define the business relationship that competing organizations need to adopt to be successful in the electronic information age. The meaning of the term is based on the concept that competing organizations need to cooperate in certain areas so that they can compete in the areas where they provide the most competitive added value. Physics publishers compete and add value with their publishing process and in the information that they publish. This should not change in the age of electronic information.

In today's paper based world, the distribution and storage medium for most physics publishers is paper. Almost all physicists can access and use this medium. Physicists today, using a set of common tools, can integrate the materials from all physics publishers into a personalized package of their choosing. Paper, in fact, is the lowest common denominator that enables physicists to integrate information from all physics publishers as well as from their note books and other information sources.

We are moving into the age where the primary distribution vehicle for information is in an electronic format. It is vital that we keep a weather eye on the physicist and insure that we, as publishers, continue to distribute information in a manner that enhances, not hampers, the physicist's ability to integrate information from diverse sources.

It is also critical that we publish in such a way that the journals can be re-used, re-packaged or re-distributed on a wide variety of media and through a wide variety of access methodologies. Most journals today are composed for paper only distribution and consumption. We have all learned that the process of composing for paper only is inadequate for any other form of distribution or access. We are all searching for a methodology that will allow us to re-use the information that we publish. Most of us are deciding that SGML is that methodology. My feeling is that composing journals under SGML will permit the physics publisher to exploit the information in all of the ways that current and future technology will allow.

SGML will clearly provide physics publishers with the capability of reusing information. There is danger, however, in each of us developing our own SGML approach. During the process of designing and evaluating The American Physical Society's RFP to publish Physical Review Letters on-line, it became apparent that one of the requirements for an electronic journal is the ability to access information from multiple publishers through a single access interface. There is danger that we will lose sight of the user, the reader, the physicist, and their need to integrate information from multiple sources.

If we each develop a unique SGML DTD, etc., we are forcing downstream agents, librarians, institutions, etc, to provide the tools and services that will enable the integration of published physics information. The downstream agent will need a unique importing approach for each publisher, just to integrate information for their users. I believe that this will add time and expense to an information dissemination process that we are all trying to streamline, improve the timeliness of and add value to.

When The American Physical Society decided to move into electronic distribution, we decided to migrate the composition process into an SGML process. We have adopted the SGML DTD defined in ISO12083 as the standard for Physical Review Letters (PRL). Beacon Graphics is providing the composition for PRL, using Arbortex. Beacon will deliver SGML files and PostScript as well as camera ready copy to the printer. We will use the SGML and PostScript files as the base for our electronic publishing deliverable. Our goal is to eventually compose all of Physical Review in a way that will provide SGML and PostScript. I anticipate that by July of 1995 we will have all of Physical Review in SGML conforming to ISO12083. The European Physical Society will also support this ISO standard.

I propose that we use this forum as a SGML sounding board for physics publishers. I propose that we cooperate in creating an SGML standard for the physics community.

Archive of Discussions on SGML Issues

APS has established a server for general discussion of issues specific to SGML and its impact on the scholarly publishing process. All postings to this server will be retained here in a hypermail archive.

This list is an informal discussion platform/work group for the airing of views on the creation and implementation of an SGML standard for use in the publishing/dissemination of physics information.

To subscribe, please send an e-mail to '' with the words 'subscribe sgml' in the body of the message. Subscribers to the sgml list will receive copies of all messages posted. Replying to a specific message subject and to the list address, '' will insure that your response is indexed properly.

To unsubscribe, please send an e-mail to '' with the words 'unsubscribe sgml' in the body of the message.

Please do not send subscribe and unsubscribe requests to ''.

All messages sent to '' will be indexed here using 'hypermail'

You should only subscribe to the list if you wish to have all postings sent to you. You may respond to any message in the list by sending an e-mail to and in the subject of your message put "Re: xxxxxxxx", where xxxxxxxx is the exact wording of the subject of the message that you are responding to. Please note that the space after the "Re:" is crucial to proper indexing of your message under the subject in the archive. The automatic reply feature of many e-mail programs does this.

To start a new discussion or topic, use a unique phrase in the subject line.