peters@cubsvax.UUCP (Peter S. Shenkin) (05/29/85)
I recently wrote a freshman chemistry study guide to accompany a new text (Segal's "Chemistry, Experiment and Theory"), and had to supply camera-ready copy to the publisher (Wiley). The text was being processed with troff, making extensive use of tbl and especially eqn, but the best output device we had was the Versatec, which wasn't good enough. So I started looking around for laser printers to typeset the final output. I looked at an Imagen 8/300, a LaserWriter and an Imagen 10 (older model, 240/in resolution; I may have the name a bit off; it may be Impress 10 or Imprint 10 or something). I ended up using the last of these because it was free, but got to look at extensive samples (Like 20-40 page chapters with equations) on the LaserWriter and the 8/300 in the course of deciding. After reading Les Earnest's and Brian Reid's interchanges, I decided to pull out what I could find of past samples and take a look. Unfortunately, I couldn't find the 8/300 sample, but my recollection is that, except for resolution, it looked pretty much like the Imagen 10 sample. I wish I had all three to look at together, so take what you are about to hear with your own grain of salt. My impression is that LaserWriter wins. The fonts on the Imagen are just too ugly. I see no problem with spacing on the LaserWriter. I realize there are many caveats, mostly involving defaults. The Imagen files were printed with default fonts, and the LaserWriter output was printed using Transcript to Postscript, which may adjust spacing better than the samples Les is looking at. But I wish I had had a LaserWriter available at the time I was running off my book. I will say that Transcript/Postscript drew crummy square-root signs (too far above the argument, and not extending low enough to the left); on the other hand, troff -t, when filtered into an Impress file using whatever program Imagen supplied, drew square root signs through the middle of the argument, so that when running off the final document square roots had to be removed from the input file, then drawn in by hand. (It's now years later than this software was supplied, so it's possible this bug has been fixed.) Incidentally, I'm sending this to fa.laser-lovers via Pnews from a Usenet site. Does anyone know if an article posted in that manner gets back to ARPA? I.e., will Les and Brian get it? Peter S. Shenkin philabs!cubsvax!peters