Consider the following file test.tex
\documentclass[11pt]{report}
\usepackage[T1]{fontenc}
\usepackage[protrusion,expansion]{microtype}
\begin{document}
\chapter{A new chapter}
Some actual testing text here. Let us make this long enough so that it
will run over at least one line so I can test to see if protrusion is
the problem. Ha Ha Ha.!!
\end{document}
If I process it with pdflatex, the file generates just fine. If I try to process it with pdflatex -output-format dvi, it fails with
kpathsea: Running mktextfm ecrm1095+20
mktextfm: Running mf-nowin -progname=mf \mode:=ljfour; mag:=1; nonstopmode; input ecrm1095+20
This is METAFONT, Version 2.718281 (Web2C 2009)
kpathsea: Running mktexmf ecrm1095+20
! I can't find file `ecrm1095+20'.
<*> ...our; mag:=1; nonstopmode; input ecrm1095+20
Please type another input file name
! Emergency stop.
<*> ...our; mag:=1; nonstopmode; input ecrm1095+20
Transcript written on mfput.log.
grep: ecrm1095+20.log: No such file or directory
mktextfm: `mf-nowin -progname=mf \mode:=ljfour; mag:=1; nonstopmode; input ecrm1095+20' failed to make ecrm1095+20.tfm.
kpathsea: Appending font creation commands to missfont.log.
! Font csnameendcsname=ecrm1095+20 at 10.95pt not loadable: Metric (TFM) file n
ot found.
If I set the option expansion=false to microtype, the DVI output compiles just fine.
Questions:
(a) Is this normal, and is there a good explanation why DVI output breaks while PDF output works? (b) Can this be fixed? or is font expansion something I just shouldn't play with in DVI output?