[tlbuild] dvi test?

[John Hawkinson]

(trimmed cc)

Paul Vojta <vojta at math.berkeley.edu> wrote on Tue, 25 Jan 2022
at 01:24:13 EST in <Ye+XjS1WXBFUlUO0 at math.berkeley.edu>:

> Hm.  I don't see anything there.

Agreed, I do not think there is any reason to suspect any kind of shared library issue, or even any meaningful X11 interaction, unless X11 is just breaking stdio or something (...). Definitely nothing like the Xt/Xaw dylib mismatch from the 2016 XQuartz bug that was segfaulting.

> I notice that the dvi file has some PostScript specials in it.
> Does the problem occur when the PS specials are removed?

That's from the \usepackage{draftcopy}. Recall the original .tex file is relatively short, it's copied below for convenience.

It is part of the "fragility" I referenced in the original bug report. It seems like removing the \usepackage{draftcopy} is sufficient to stop the problem, but that also seemed to be true of so many other things in the original .tex file that I'm just not sure anymore. Indeed, when I first started to debug this with my original source file, there was a \LaTeX macro invokation in what is now the lorem ipsum text, and when I changed it to the word "latex" everything seemed fine, which had me scratching my head for quite a while. But then it became clear that the problem was still present, just far less consistently, appearing on the 30th or 50th press of shift-R.

That said, as I stand here right now on 5 hours sleep, removing the \usepackage{draftcopy} from "t2a" appears to resolve the issue, so it may well be a required dependency. But my memory is that removing it was one of the first things I tried (unsucessfully) in narrowing the problem, so I don't trust that conclusion.

> If you wait for the ghostscript process to fully start up before pressing 'R'
> (with the original dvi file), does the problem still occur?

Yes it does still occur. I don't think there is any timing-related aspect here, you can wait 1, 5, 10, 30 seconds before pressing shift-R and it will crash the same way on the first try. But the idea that this has to do with xdvi's subprocess management is intriguing.

Karl Berry suggested invoking xdvi with "-debug -1".
Although that doesn't produce any obvious hints, it does produce a lot of output.
When I have some more time later today I will try and compare that tracing between the working and the non-working machines and see if any differences are instructive.

(but for now, more sleep in the offing)

--
jhawk at alum.mit.edu
John Hawkinson


% begin t2a.tex
\documentclass{article}
\usepackage{draftcopy}

\begin{document}
\section{Table of Authorities}
\sc a\hfill\rm B\par
\newpage
xxx
\newpage
\begin{center}\textbf{CERTIFICATE OF TYPE-VOLUME COMPLIANCE\\
    and COLOPHON}\end{center}


Lorem ipsum dolor sit amet, consectetur adipisicing elit, sed do eiusmod
tempor incididunt ut labore et dolore magna aliqua. Ut enim ad minim veniam,
quis nostrud exercitation ullamco laboris nisi ut aliquip ex ea commodo
consequat. Duis aute irure dolor in reprehenderit in voluptate velit esse
cillum dolore eu fugiat nulla pariatur. Excepteur sint occaecat cupidatat non
proident, sunt in culpa qui officia deserunt mollit anim id est laborum.

Lorem ipsum dolor sit amet, consectetur adipisicing elit, sed do eiusmod
tempor incididunt ut labore et dolore magna aliqua. Ut enim ad minim veniam,
quis nostrud exercitation ullamco laboris nisi ut aliquip ex ea commodo

\end{document}