tlug: Simple Ghostview Question

["Stephen J. Turnbull" <turnbull@example.com>]

>>>>> "Rob" == bickel  <bickel@example.com> writes:

    Rob> I have ghostview 3.58 that I installed from PJE-0.1beta (the

Huh?  AFAIK ghostview is still at 1.5 or something like that; I have
not looked at it for a while, but that's where it was say 6-8 months
ago.  It hasn't been actively maintained for quite a while.

GNU Ghostscript was at 3.53 last I checked, that was about 3 months
ago.  Lessee....

Hm.  At the official Ghostscript distribution site (ftp.cs.wisc.edu), 
we have under ghost/gnu

ghostscript @ 3.33; I would guess 3.5x will be out soon.
ghostview   @ 1.5
gv          @ 3.5.8; this is a separate development branch, but I
                     guess it's OK to call it ghostview since that's
                     the origin and the progenitor appears defunct

Recent Ghostscripts are under ghost/aladdin.

    Rob> new Japanese packages).  This is the only program for which I
    Rob> can not see the Japanese correctly.  If I have a tex file, I
    Rob> can see the Japanese correctly as I write it in mule or vi on
    Rob> kterm.  When I tex it, I can see the Japanese correctly in
    Rob> the .dvi file with xdvi.  When I convert the .dvi file to a

Then you're almost surely using a Japan-localized version of xdvi
which uses virtual* fonts.  (Virtual* because they are not the
official notion of virtual font which is used by TeX and dvi*k
programs.)  Which version of Japanese TeX are you using?

    Rob> postscript file with dvi2ps, I can not see the Japanese

Try dvipsj instead.  dvi2ps was always pretty cranky in my
experience.  That was another problem I had with JE: they invariably
used the least compatible localizations of common programs.  Looks
like they're still at it.

    Rob> correctly with gv (ghostview) or gs (ghostscript).  When I
    Rob> print it out, it looks fine, Japanese okay.  Now also, when
    Rob> one of my students rlogins in to my machine and uses my gv or
    Rob> gs, but displaying on his machine, he can see the Japanese.
    Rob> When he logs in directly to my machine, he can not.  I must

This sounds highly unlikely to me, but I'll take your word for it.
Get the student to do `echo $PATH' in both situations.  Look for
multiple versions of Ghostscript (it's probably not a problem with gv
or ghostview) on the PATH.  What is the ghostscript console saying
while this is happening?

    Rob> then assume the problem will be somewhere in how I display
    Rob> Japanese, but again on any other program, I can see Japanese
    Rob> with no problem(this includes xdvi, kterm, mule, netscape,
    Rob> etc).

xdvi uses .pk fonts.  If necessary, it will use MKTEXPK to build them
at run time from Metafont sources.  All of the others use X fonts.
Ghostscript uses (surprise, surprise) Postscript fonts, except for the
hacked-up vfonts versions and the platform fonts version.  (Note that
as far as I know the vfont versions of Ghostscript do not use the
platform fonts interface---they patch the source :-(.  If you can
check it out, let me know; I'd like to find out I'm wrong.)

    Rob> Does anyone have an idea of where I should look to determine
    Rob> what is wrong here?

There is an option to use `platform fonts' with Ghostscript; in this
case it would allow you to use X.  It's possible that the PJE guys
decided to build that in to your version of Ghostscript.  do a `gs
--help 2>&1 | less' and find out what version you're using, and what
options are in it.  what the hell, while you're at it, do `( gs --help
; gs --config ) 2>&1 | sendmail turnbull@example.com'.

However, it seems rather unlikely that this would be done.  It's even
more unlikely that TeX would produce a file that invoked X fonts,
which are pretty ugly when printed out.  Most probably you have
Japan-localized versions of TeX and Ghostscript which use virtual*
fonts, and the problem is in the dvips.

Another possibility is that by default ghostscript searches for all
kinds of stuff in `.' first, hardwired, which is a very, very bad
idea.  Lots of people have argued with Peter about this, but he seems
unwilling to rethink the interface so that users can search for their
own files there if they want without also picking up init files and
library files there.  However, a compile-time switch is provided to
alter this behavior, and the PJE guys may have decided to do that.  In 
which case it is likely that ghostscript will display erratic
behavior.

Next TLUG meeting is Saturday Dec. 13, 1997  (possibly Nov. 13?)
---------------------------------------------------------------
a word from the sponsor:
TWICS - Japan's First Public-Access Internet System
www.twics.com  info@example.com  Tel:03-3351-5977  Fax:03-3353-6096