Path: utzoo!attcan!utgpu!jarvis.csri.toronto.edu!mailrus!accuvax.nwu.edu!delta.eecs.nwu.edu!phil From: phil@delta.eecs.nwu.edu (William LeFebvre) Newsgroups: comp.text Subject: Re: missing postamble/iptex Keywords: help! Message-ID: <1100@accuvax.nwu.edu> Date: 29 Aug 89 15:49:11 GMT References: <2932@tahoe.unr.edu> <19309@mimsy.UUCP> Sender: news@accuvax.nwu.edu Reply-To: phil@delta.eecs.nwu.edu (William LeFebvre) Distribution: usa Organization: Northwestern U, Evanston IL, USA Lines: 27 In article <19309@mimsy.UUCP> chris@mimsy.UUCP (Chris Torek) writes: >In article <2932@tahoe.unr.edu> mikew@wheeler.wrcr.unr.edu (Mike Whitbeck) >writes: >>I got a version of TeX/LaTeX (HaTe ThIs!) going on an Atari ST .... >>However when I send the dvi file to my friendly SUN 3 at work >>the printer (iptex, shell script calling an IMAGEN driver, >>imagen1) program complains about a missing "postamble" and >>questions if this is REALLY a dvi file then quits. > >The first step in examining a questionable dvi file is to run dvitype >over it. If dvitype likes it, there may well be a bug in the dvi >interpreter (in this case, probably an ancient version of imagen1). I remember a bug in older versions of iptex whose symptoms are similar to this. It would not tolerate more than a small number of filler bytes at the end of the DVI file (16 comes to mind). If it didn't find the end of the postpost within a short distance of the end of the file, it complained that it couldn't find the postamble. We discovered this when moving a dvi file made under VMS over to a Unix machine. The VMS TeX padded out with the dvi filler byte to the end of the last block. This is perfectly legal dvi, but at least one older version of iptex didn't like it. William LeFebvre Department of Electrical Engineering and Computer Science Northwestern University