mrwittma@phoenix.Princeton.EDU (Martin R. Wittmann) (03/11/89)
/* I recently wrote this letter to Nelson Beebe at Utah and thought I * would post it for general enlightenment. Any solutions floating around * out there??? * * */ Dear Mr. Beebe: Please pardon my fairly lengthy note. It concerns truncation of the right side of a page by DVIEPS, and problems with loadchar(). Some time ago I wrote you about difficulty using dvieps: > 1) I have a Panasonic 1091i printer which is supposed to look just like an > RX-80 (I think) in "standard mode." Yet when I tried to print a small > ".eps" file (a letter to dad), I got the first line of the return address, > the date, "Dear Dad", some interspersed codes and several formfeeds, and > nothing else.... I got this helpful reply: ] From: ftg@gatech.edu (Gary Peterson) ] The fix, from Gary Beihl the author of DosTeX, is to use: ] copy file.eps /b prn I felt pretty silly! It works. (Actually, I use: copy /b file.eps lpt1) But, "ftg" and I share some common difficulties. We both have two DVIEPS programs, which we will call: D-DVIEPS from the DosTeX archives on wsmr-simtel20; U-DVIEPS from dviexe.arc on science.utah.ed: ] But I still have problems with DVIEPS programs. ] DoSTeX's DVIEPS (D-DVIEPS) is crashing on relatively simple files. I found that it this program crashed even on very simple files. The error I got repeatedly was: FATAL--loadchar(): Could not allocate 20 words of raster space--used 843 words so far. This seemed to take place on a line with a few more capital letters than most!. Is this simply a problem of the stack being too small? (D-DVIEPS is 75348 bytes long; U-DVIEPS is 107764 bytes!) I downloaded all the back issues of the dvidriver newsletters and found a few references to stack size problems but nothing more direct. Would the utility EXEMOD help? I don't know where to find this...it is not standard DOS. The newsletter said this would resize an .EXE file's stack, without changing anything else. BTW, what do you know about this version? The documentation says it uses two pass processing of the page, but it seemed to be running more slowly than U-DVIEPS. ] Utah's DVIEPS (U-DVIEPS) is truncating everything about ] 6" from left edge of paper. I ran U-DVIEPS with the "-d56" option, and observed from the diagnostic output that setchar() thinks any character more than 5.75 inches from the left edge or about 8 inches from the top is off the current page. What defines the current page? How can this be remedied? I also ran U-DVIE72 with "-m1200 -d56"... the entire text of my letter was considered "off page." I will be very grateful for your reply or reference to other sources which may be able to resolve these difficulties. Thank you. -- martin wittmann princeton u., chem. eng.