This document was almost entirely generated using the LaTeX2HTML translator Version 95.1 (Fri Jan 20 1995) Copyright © 1993, 1994, Nikos Drakos, Computer Based Learning Unit, University of Leeds.
In the colophon, written for the book prior to starting work on the electronic proceedings, I mentioned that the reason for producing the book from a single gigantic LaTeX file was the hope that it would simplify the production of the electronic version. I am no longer sure whether this was a good idea or not. Much time went into a LaTeX2HTML style for the proceedings, which resulted in a style for AASTeX, and that will be useful. Much time went into fixing bugs in LaTeX2HTML, like the one that broke cross-references between articles. I also modified it to make the navigation buttons refer to an absolute URL rather than a relative one, so they still work on a contribution that you discover via a WAIS search. The translation itself took about 6 hours of Sun SPARC 20 cpu time, with a peak swap space use in excess of 300 MB.
The output from LaTeX2HTML was then edited by hand. Steps included renaming the files, adding the "slug" line, indentifying the volume and its editors, to each file, creating an author index, fleshing out the table of contents, and fixing the little problems. LaTeX2HTML creates files with names like node97.html. I renamed contributions to the first author's last name and first initial (or two initials in the case of hyphenated first names) followed by .html, the convention established for the previous electronic proceedings. This convention made it very easy to add links for references to contributions in the previous electronic proceedings, and should be maintained.
The previous electronic proceedings experimented with an author index where the visible text for the hot link to each contribution is the page number from the printed proceedings. Among other uses, this allows readers to derive a citation to the printed proceedings from the electronic version alone. I created such an index for these proceedings, starting from the author index to the printed version, and modified by a perl script that reads a table matching page numbers to file names.
In the LaTeX file that contains the entire book, each contribution is a chapter. From this, LaTeX2HTML created a table of contents listing all of the titles but none of the authors. I cut and pasted author lists from the table of contents to the printed proceedings. LaTeX2HTML created a separate table of contents for each part of the book, with a brief table of contents listing the part names. Being unsure whether this was better than a comprehensive table of contents, listing the title and authors for each contribution, I did both.
Some of the little projects involved playing with the conference photograph, scanned by Zolt Levay, and fixing some weird things LaTeX2HTML did with enumerated lists, some of the references, and places where it seemed to mis-match various begin and end environment markers. LaTeX2HTML inexplicably ran two parts of the book together, producing only one table of contents when there should have been two, and one of the big PostScript figures failed to be converted to an inline image.
Having made too many hand edits to run LaTeX2HTML again, I will just mention that I should have allowed LaTeX2HTML to break up the individual contributions, putting each section in a separate file. Some contributions, especially some of the invited talks, will take a long time to download as they are now.
Some authors made use of the htmllink command to enhance the translation of their manuscript into hypertext. The overall look and feel of these proceedings is still pretty "flat", however, revealing that it was translated from another medium rather than being a native work in hypertext. On the other hand, these proceedings illustrate some of the shortcomings of (the current version of) the Hypertext Markup Language when it comes to typesetting scientific manuscripts: the inability to do tables, and the lack of a sufficient set of special symbols (see Karen Strom's paper, ) or a way to do equations.
I would like to thank Jeannette Barnes and Rudi Albrecht for their comments on the first draft of the electronic proceedings, which was basically the raw output from LaTeX2HTML. The final product was greatly improved as a result of their suggestions. I also want to thank Daniel Durand, Jeannette Barnes, and Dennis Crabtree for their work on the first electronic proceedings. They are directly responsible for the current undertaking, by showing that it could be done. In details that should be apparent, I also learned how it should be done by examining their work.