[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: drift zeroing and dvicopy
- To: ls@mathp7.jussieu.fr (Larry Siebenmann)
- Subject: Re: drift zeroing and dvicopy
- From: Petr Sojka <sojka@daeron.ics.muni.cz>
- Date: Fri, 22 Apr 1994 13:13:54 +0200 (MET DST)
- Cc: math-font-discuss@cogs.susx.ac.uk
- Content-Length: 1717
- Organization: Masaryk University, Brno, The Czech Republic
- Postal-Address: Institute of Computer Science, Buresova 20, 60200 Brno
- Telephone: +42-5-41321237 ext. 283, Fax: +42-5-41212747
: LS=me
: AT=Andy Trevorrow, author of OzTeXandrew.trevorrow@anu.edu
:
: LS> Do you believe the .dvi's (expanded)
: LS> that Oz produces are immune to Jiri's
: LS> problem? [ie the drift zeroing for dvicopy output]
:
: AT> I've certainly never seen any such problems, but
: AT> then I don't use a lot of accents in the stuff I write!
: AT> Nobody has ever pointed out any problems in this area yet
: AT> and some people are obviously using vfs quite heavily
: AT> now.
:
: In the present circumstances where existing literature
: seems to provide inadequate support for a theoretical
: discussion, might I suggest that readers report
: any *empirical* evidence of quality degradation.
I may report empirical evidence of quality degradation
and some possible remedy too:
I use many accents in Czech fonts so I noticed
misplacement of accents in virtual fonts:
HP LJ III+PS cartridge, 300dpi, dvips 5.49, no extra switches
Using -E0 itself doesn't help, BUT
with the exactly the same configuration, .dvi and dvips
using -D1000 -E0 solved the problem:
dvips used more precise units for positioning, no drift allowed
[and I had no bitmap fonts in document], printer probably ignored
high resolution setting, but positioning worked well.
The .ps file generated by dvips in the second case
was slightly bigger (<5%), but who worry.
Having new switch for setting of dvips `positioning units'
or resolution would solve the problems (or rather
having some better dvitype like standard for virtual font
handling [any people from dvi standard committee here?]).
Petr Sojka
P.S.
: The most probable teltale sign might be horizontal
: misplacement of pieces of virtual characters.
Yes, it was the case.