I am writing a new textbook entitled, ``EARTH SOUNDINGS ANALYSIS: Processing versus Inversion'' (PVI) and I am revitalizing my 1985 textbook, ``Imaging the Earth's Interior''(IEI). Both books can now be viewed on a workstation screen, including all the figures. The new book should be publishable by the end of the year.
Nearly all the figures in PVI have underlying program support. Of 140 figures, 40 are supported by interactive programs, and about 70 are supported by batch programs, most of which run in a matter of seconds. The few figures without program support are mostly drawings and indications of ideas rather than demonstrations of them.
The documents themselves are written in LATEX which can be printed on a laser printer, or the page images can be displayed on a workstation using Dirk Grunwald's xtex (freely available via anonymous ftp from foobar.colorado.edu 188.8.131.52) that enables us to include pushbuttons in the manuscript. Steve Cole adapted this software to X on Sun workstations so I am delighted to be seeing bitmap pictures of pages on the screen of my workstation, including pasted up figures. With xtex we can place a pushbutton anywhere in the text (see Karrenbach and Nichols, 1990, page 150) and the pushbutton will launch any command in the system. This provides more flexibility than we can begin to exercise, so I began (with help of Martin) by having each figure caption provide a pointer to the figure building file.
All the figures in PVI now contain pushbuttons, which either launch the interactive programs in an appropriate initial state, or for the batch programs, check to see whether the plot file is up-to-date with the software that generates it; then if not, they recompute it; and finally they invoke the local plot display program which has color and zoom. For 23 figures, the ed1D interactive program [Claerbout, 1988] is invoked, and for 16 figures the Zplane program [Claerbout, 1987] is invoked. We have not fully defined the kind of ``workbench'' that should surround batch-program figures. Although we want to offer the reader some flexibility, it is not clear how much. Obviously, maximum flexibility comes from leaving the manuscript and logging in at the figure construction directory. The beginning of a possible ``workbench'' environment is suggested by Steve Cole's Xvpanel [Cole, 1990]; but more experimentation is needed before installing it in my books.
For the older book, IEI, the program support is in place for 30 of 170 figures For those 30, pushbuttons will be installed soon. The rest of the figures are in a raster scan format. About 30 more figures should get program support over the next year. Many of the scanned figures are sketches not warranting an underlying program. About 40 of these sketches are being being redrawn with a simple draw-type program to escape from the crudeness of scanning. Others contain hyperbolas or other elements that are beyond simple draw-type programs. Still other figures deserve program support, but providing it may be beyond my time and energy. The scanned figures now occupy about 100Mb, which is more than ten times everything else put together. Luckily, the compress program works well. Of the two books, IEI is more suitable for movies than PVI, but PVI is more suitable for interaction. With next quarter's teaching being from IEI, some figures should turn into movies.
I have devoted much of my time to cleaning up the books themselves and the underlying figure makefiles (actually cakefiles) and little of my time to designing ways of making the final materials more versatile and appealing to others. I hope to find a partner interested in the design of the ``workbench''.