Xref: utzoo comp.text.tex:4699 comp.fonts:1801
Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!yale!mintaka!wald
From: wald@theory.lcs.mit.edu (David Wald)
Newsgroups: comp.text.tex,comp.fonts
Subject: Re: mf2ps
Keywords: postscript, mf2ps
Message-ID: <1991Jan8.200230.7139@mintaka.lcs.mit.edu>
Date: 8 Jan 91 20:02:30 GMT
References: <1991Jan8.055007.17766@midway.uchicago.edu>
Sender: daemon@mintaka.lcs.mit.edu (Lucifer Maleficius)
Reply-To: wald@theory.lcs.mit.edu (David Wald)
Followup-To: comp.text.tex
Organization: Laboratory for Computer Science, MIT
Lines: 44

In <1991Jan8.055007.17766@midway.uchicago.edu> amun@ellis.uchicago.edu
(james frederick amundson) writes:
>
>OK, I found mf2ps.  (For those of you who didn't read about it
>earlier, mf2ps creates postscript outline font versions of metafont
>fonts.)  After having read the article, I'm sold on it.

Having just read the article, I am unconvinced.  The authors' goal, in
an abbreviated form, is to automatically take METAFONT descriptions
and produce PostScript outline fonts of equal quality and smaller
size.  The fonts shown in the article had neither.

Yanai and Berry did most of their font comparisons at 300dpi.  This
was a mistake.  At that resolution, the PostScript Computer Modern is
from three to seven times the size of the METAFONT-produced bitmap
font, depending on the magstep.  Only if your document demands all of
magsteps 0, .5, 1, 2, 3, and 4 do you get even a slight space savings.

As for quality, the results given are somewhere between useless and
dishonest.  The PostScript fonts at 300dpi produce results with
several inaccuracies and, in general, a much rougher look than the
METAFONT fonts.  After this is demonstrated, the authors go on to
compare the PostScript fonts at 600dpi and 1270dpi with those same
300dpi METAFONT fonts.  This is absurd.  Nor does it help that, even
at 1270dpi, the PostScript fonts contain some of the same inaccuracies
they did at 300dpi.  It was humorous to read, however, that in
peforming this unfair comparison, they performed all paste-up by hand
"for fairness in the comparison".

The work they've done is, in fact, somewhat interesting, and it could
eventually be useful for high resolution printing.  However, it is
also very preliminary.  (Also, it doesn't increase my confidence in a
font-hacking article when the fonts used in setting the article itself
are frequently rough and broken.)  As it stands, I don't have enough
reason to try dealing with their Pascal code.  I would, however, be
interested in hearing about any success people have with it,
especially at high resolutions.

-David
--
============================================================================
David Wald                                           wald@theory.lcs.mit.edu
   "Blessed are the peacocks, for they shall be called sonship of God"
             -- Matt 5:9, from a faulty QuickVerse 2.0