Automatica, March, 1996, Volume 32, No. 3
Scientists and engineers, like many other professionals, can hardly get together without text processing finding its way into the conversation. A few "illiterates" left alone, everyone has an opinion and preferences when it comes to particular text processors or text processing systems.
Computer scientists, mathematicians, and theoretical engineers, by and large, appear to favor LaTeX, and use it on workstations. For many practical engineers, LaTeX is no option because they useWhat You See is What You Get text processors on PCs.
The pie diagram of Fig.1 confirms this observation.
Fig. 1. LaTeX and WYSIWYG documents at the 1995 European Control Conference (ECC) and the 1995 IFAC Conference on Intelligent Autonomous Vehicles (IAV).
The data for this diagram were obtained by scanning one volume of the preprints of the 1995 European Control Conference and the preprints of the 1995 IFAC Conference on Intelligent Autonomous Vehicles. The ECC is dominated by theoretical engineers and applied mathematicians. More than two thirds of them use LaTeX. IAVs largely are in the domain of engineers who think in concrete terms. Three quarters of the papers at their conference have a WYSIWYG look.
LaTeX is favorite with the theoreticians because it provided high quality mathematical typesetting long before WYSIWYG text processors did. The situation has changed, though. All leading WYSIWYG programs currently provide for mathematics. Figure 2 shows the same formula produced with LaTeX and with a leading WYSIWYG text processing program. The formulas were printed with the same 600 dpi PostScript printer. LaTeX is better at spacing, positioning and proportioning. It is clear from the figure that the quality gap is no longer large, though. The figure also shows that inputting maths in LaTeX is not for everyone. Several but not all WYSIWYG word processors, and one LaTeX editor, offer an attractive user-friendly graphical interface based on templates.
Fig. 2. (a) LaTeX formula. (b) WYSIWYG formula. (c) The LaTeX code.
Scientific books and journals mostly still are typeset by rekeying text from hardcopy manuscripts. With the multitude of electronic documents that are now being generated it is logical to try to eliminate the rekeying. Publishers and printers have been slow in doing this. An important reason is that typesetters' files are strongly dependent on the typesetting system and conversion from other types of text processing systems involves a prohibitive amount of editing.
AUTOMATICA's publisher Elsevier Science and other publishing companies are now trying to break through this barrier. Since the beginning of 1995 Elsevier sollicits electronic scripts for its journals, including AUTOMATICA.
With 600 dpi laser printers and LaTeX, camera ready copy may be produced that is good enough to be included in journals. Rekeying is avoided if authors submit the final versions of their papers as LaTeX source files. At this stage Elsevier uses a single generic LaTeX style file, called elsart, for all its journals. The hardcopy this style file produces does not imitate any particular journal. When the files delivered by the author reach Elsevier they are copy edited and printed with the help of a special style file for each journal. Among other things this adds the logo and catch lines and switches to Monotype Times and the right page size. The result is camera ready hard copy in the appropriate style that is merged with what comes from the typesetter.
The situation is less fortunate for WYSIWYG documents. Elsevier accepts the files, converts them to a single leading standard WYSIWYG format, edits them and sends them to the typesetters who try to use them. If anything goes wrong the paper is rekeyed.
Eventually, in a year or so, Elsevier plans to go a full Computer Aided Production (CAP) system. Then everything will be converted to Standard Generalized Markup Language (SGML) documents that conform to a single Document Type Definition (DTD). The typesetters work directly from these without any rekeying, and the journal style is a typesetting function. All figures are electronic, scanned from hard copy originals if necessary.
For authors submitting electronic scripts has the advantage that because there is no typesetting the printed paper has no errors except their own and that -- eventually -- production time will be reduced. Moreover, the document may be included in electronic archives this way. The Instructions for Contributors to AUTOMATICA on the inside back cover provide guidance how to prepare and submit LaTeX scripts.
An entirely different question is whether traditional hardcopy scientific journals will survive the advent of the electronic publishing era. Nobody, seems to know, not even publishers. Time will show.
I am grateful to Sebastian Rahtz of Elsevier Science for supplying some of the factual information for this editorial.
Huibert Kwakernaak, Editor-in-ChiefLast modified on November 11, 1995