Current Shareware 1994 January/SHAR194.ISO/misc/linhowto.zip/PRINTING.HOW

Text File | 1993-10-08 | 90KB | 2,185 lines

Newsgroups: comp.os.linux.announce,comp.os.linux.admin
From: [e-mail] (Grant Taylor)
Subject: Linux Printing HOWTO
Keywords: Linux, printing, lpr
Summary: HOWTO on printing under Linux
Followup-To: poster
Approved: [e-mail] (Matt Welsh)

Archive-name: linux/howto/printing
Last-Modified: 28 August 1993

Linux Printing HOWTO
By Grant Taylor <[e-mail]>
and Brian McCauley <[e-mail]>
------------------------------------------------------------------------------

This is the Printing HOWTO for Linux.  This is a member of the second
generation of Linux FAQS.  The original Linux FAQ gradually became a
monolithic beast, and has now been re-written in a new syle, refering
to HOWTOs for details.  This HOWTO details how to properly set up most
types of printers, how to configure the software to make them print
well, what types of fax software are available, and how to preview
many types of printable output.  Other HOWTOs detail networking,
kernel hacking, etc.  All HOWTOs are found in several places, the
official location being sunsite.unc.edu:/pub/Linux/HOWTO, and another
being rtfm.mit.edu.

Topics:
         INTRO
         Summary of changes since release aug 05-93
         PRINTING SOFTWARE   <<<<<<<-----------
         ascii
         pr
         postscript
         fonts
         faxing
         *roff/man pages
         mpage
         a2ps
         enscript
         gslp
         psutils
         tex/dvi
         dvilj
         texinfo
         hardware and kernel drivers
     Printer device names
         LPR                <<<<<<------------
         What lpr and lpd do
         Where do the files go and what should their permissions be?
         What are all these diretories for?
         lpd not working
         Where do I get a printcap for a xxxxx?
         The semantics of /etc/printcap
         The syntax  of a /etc/printcap
         An /etc/printcap gotcha.
         The minimum /etc/printcap?
         How do I prevent the `staircase effect'?
         Resetting the printer after each printout
         Preventing formfeed after each printout
         Printing to a serial port
         Serial settings in printcap
         cat works to the serial port, but not lpd (1)
         cat works to the serial port, but not lpd (2)
         Printers that aren't simple character devices
         Getting burst/banner pages
         Printing text on a PS printer
         Trunation of graphics files
         Why doesn't `lpr -i' work?
         Why doesn't `lpr -p' work?
         lprm says `too many users'
         lpc complains that it hasn't stopped a daemon
         lpr over a network
         Writing lpd filters
         A filter tests OK but fails when lpd uses it
         `of' filters.
         Getting filters for given printers
         Filters from programs that won't read STDIN
         Having many filters
         What are magic filters?
         Magic filter examples
         PREVIEWING          <<<<<<------------
         ghostview
         gspreview
         xdvi
         xtex
         gxditview
         non-x-previewing
         ASCII TRANSLATION   <<<<<<------------
         from TeX
         from dvi
         from ps
         from troff
         from ascii
         CONTRIBUTORS        <<<<<<------------

--------------------------------------------------------------------

Subject: INTRO
From: [e-mail]

This is the Print-HOWTO for Linux.  Please send comments, bug reports,
etc, to:

[e-mail]  (Grant Taylor) or
[e-mail] (Brian McCauley)

release Sep 12-93

This is a joint effort; originally, we both produced separate
documents - the lpd-HOWTO, and the printing.how-to, but now we have
joined forces.  The old lpd-HOWTO, which has been independent and a part
of the main Linux HOWTO, now makes up the LPR section of this document.
No non-digest version of either the printing.how-to or the lpd-HOWTO is
current (although most of the information is correct).

If you undigestify this document with your mail reader, you can simply
hit reply, and the reply will go to the maintainer of each section.

If you know anything or can point me towards any useful programs not
in this document that should be, please let me know!

I have a mail server set up on my machine.  It stores documents and
other things related to printing and previewing with Linux, including
the latest and greatest version of this document.  Send mail with the
subject `help' to:

[e-mail]

for further information.  (This address only valid until Spetember.)
The new address will be announced when it works.

------------------------------

Subject: Summary of changes since release aug 18-93
From: [e-mail]

now a HOWTO        -  we have joined forces with mdw and now fit
               into the greater scheme of things as a part
               of the FAQ/HOWTO project.

Many small corrections...

Important notice!!  The ftp site (and whole domain of)
export.lcs.mit.edu will be changing it's name, and eventually number,
to ftp.x.org.  This is a result of the separation of the X Consortium
from the MIT Computer Science lab.  This change becomes officially
effective August 18 (my birthday!), but both names will work already
and for a while to come.

------------------------------

Subject: PRINTING SOFTWARE   <<<<<<<-----------
From: [e-mail]

These sections describe printing software known to be available for
Linux.  Note that most Un*x printing software can be easily ported to
Linux.

------------------------------

Subject: text
From: [e-mail]

Under Linux, or under most any Un*x operating system, the easiest
thing to print with most printers is plain ASCII.  Any way you can
arrange to send the text to the printer is perfectly valid.  If you
have a serial printer, then try devices /dev/ttys?, /dev/ttyS?,
/dev/cua?, etc, and if you have a regular parallel hookup, use
/dev/lp?.  typing

cat file >/dev/????

should do it.  You may need root privileges, or you may wish to chmod
your printer device if this is all you ever print.  (Note security
hole there if you've more than one user)  Some printers (ie, HP
DeskJet) want dos-style end of lines: newline+carriage return.
These printers will show the 'staircase effect'; further information
on how to deal with it is in the section "How do I prevent the
staircase effect."

------------------------------

Subject: pr
From: [e-mail]

Most plain ascii files in the unix world have a tendency to be just
that - plain, unformatted ascii with no page breaks or anything else
to make a printed copy look nice and not have lines printed over
perferations.  The answer to this problem is to run your text through a
formatter such as pr.  Pr is a standard unix utility designed to
format plain text for printing on a line printer.  The usual
appearance of the resulting formatted text has a header and/or footer,
page numbers, the date, possibly margins, double spacing, etc.  As is
common with unix utils, pr has a bazillion options.  They are detailed
in the man page (which you should read), but a few useful ones are:

       -d     Double space the output.

       -e[in-tab-char[in-tab-width]]
              Expand  tabs to spaces on input.  Optional argument
              in-tab-char is the  input tab  character,  default
              tab.   Optional  argument in-tab-width is the input
              tab character's width, default 8.

       -h header
              Replace  the filename in the header with the string
              header.

       -l page-length
              Set  the  page  length  to  page-length lines.  The
              default is 66.  If page-length is less than 10, the
              headers  and  footers  are  omitted,  as  if the -t
              option had been given.

       -n[number-separator[digits]]
              Precede each column with a line number; with paral-
              lel  files,  precede  each line with a line number.
              Optional argument number-separator is the character
              to  print after each number, default tab. Optional
              argument digits is the number of  digits  per  line
              number, default 5.

       -o left-margin
              Offset  each  line with a margin left-margin spaces
              wide.  The total page width is this offset plus the
              width set with the -w option.

       -w page-width
              Set the page  width  to  page-width  columns.   The
              default is 72.

------------------------------

Subject: postscript
From: [e-mail]

Printing almost anything other than plain text under unix involves
the ability to print postscript.  If you have a postscript printer,
you're all set.  But for most, this is not so easy.

The established way to interpret postscript under Linux is to use
ghostscript, which, like nearly everything else, comes from the GNU
project.  Ghostscript is a postscript interpreter which accepts
postscript input and generates output appropriate for X displays, most
printers, some specialized display hardware, and fax software.

The following devices are available as of ghostscript 2.6.1pl4:

        linux   PC vga using linux vgalib
#       x11     X Windows version 11, release >=4   [Unix and VMS only]
# Printers:
# *     appledmp  Apple Dot Matrix Printer (should also work with Imagewriter)
#       bj10e   Canon BubbleJet BJ10e
# *     bj200   Canon BubbleJet BJ200
# *     cdeskjet  H-P DeskJet 500C with 1 bit/pixel color
# *     cdjcolor  H-P DeskJet 500C with 24 bit/pixel color and
#               high-quality color (Floyd-Steinberg) dithering
# *     cdjmono  H-P DeskJet 500C printing black only
# *     cdj500  H-P DeskJet 500C (same as cdjcolor)
# *     cdj550  H-P DeskJet 550C
# *     declj250  alternate DEC LJ250 driver
# +     deskjet  H-P DeskJet and DeskJet Plus
# *     dfaxhigh  DigiBoard, Inc.'s DigiFAX software format (high resolution)
# *     dfaxlow  DigiFAX low (normal) resolution
#       djet500  H-P DeskJet 500
# *     djet500c  H-P DeskJet 500C
#       epson   Epson-compatible dot matrix printers (9- or 24-pin)
# +     eps9high  Epson-compatible 9-pin, interleaved lines
#               (triple resolution)
# *     epsonc  Epson LQ-2550 and Fujitsu 3400/2400/1200 color printers
# *     escp2   Epson ESC/P 2 language printers, including Stylus 800
# +     ibmpro  IBM 9-pin Proprinter
# *     jetp3852  IBM Jetprinter ink-jet color printer (Model #3852)
# +     laserjet  H-P LaserJet
# *     la50    DEC LA50 printer
# *     la75    DEC LA75 printer
# *     lbp8    Canon LBP-8II laser printer
# *     ln03    DEC LN03 printer
# *     lj250   DEC LJ250 Companion color printer
# +     ljet2p  H-P LaserJet IId/IIp/III* with TIFF compression
# +     ljet3   H-P LaserJet III* with Delta Row compression
# +     ljet4   H-P LaserJet 4 (defaults to 600 dpi)
# +     ljetplus  H-P LaserJet Plus
# *     m8510   C.Itoh M8510 printer
# *     necp6   NEC P6/P6+/P60 printers at 360 x 360 DPI resolution
# *     nwp533  Sony Microsystems NWP533 laser printer   [Sony only]
# *     oki182  Okidata MicroLine 182
#       paintjet  H-P PaintJet color printer
# *     pj      alternate PaintJet XL driver
# *     pjxl    H-P PaintJet XL color printer
# *     pjxl300  H-P PaintJet XL300 color printer
# *     r4081   Ricoh 4081 laser printer
# *     sparc   SPARCprinter
# *     t4693d2  Tektronix 4693d color printer, 2 bits per R/G/B component
# *     t4693d4  Tektronix 4693d color printer, 4 bits per R/G/B component
# *     t4693d8  Tektronix 4693d color printer, 8 bits per R/G/B component
# *     tek4696  Tektronix 4695/4696 inkjet plotter
#%*     trufax  TruFax facsimile driver  [Unix only]
# File formats and others:
#       bit     A plain "bit bucket" device
#       bmpmono Monochrome MS Windows .BMP file format
#       bmp16   4-bit (EGA/VGA) .BMP file format
#       bmp256  8-bit (256-color) .BMP file format
#       bmp16m  24-bit .BMP file format
#       gifmono Monochrome GIF file format
#       gif8    8-bit color GIF file format
#       pcxmono Monochrome PCX file format
#       pcxgray 8-bit gray scale PCX file format
#       pcx16   Older color PCX file format (EGA/VGA, 16-color)
#       pcx256  Newer color PCX file format (256-color)
#       pbm     Portable Bitmap (plain format)
#       pbmraw  Portable Bitmap (raw format)
#       pgm     Portable Graymap (plain format)
#       pgmraw  Portable Graymap (raw format)
#       ppm     Portable Pixmap (plain format)
#       ppmraw  Portable Pixmap (raw format)
# *     tiffg3  TIFF/F (G3 fax)

  Lines beginning with a `#' are drivers included with the gs261
  source distribution.

  Drivers marked with % require commercial software to work

Installations of ghostscript are available from several places:

A full, and presumably correct, installation is available in the SLS;
note that this binary requires libX.so.???, as it includes the X11
display driver.

The `official' installation is to obtain the sources and build it
yourself:

prep.ai.mit.edu(or any gnu mirror):/pub/gnu/ghostscript-xxxx.tar.gz
and                                         ghostscript-fonts-xxxx.tar.gz

This is probably best, as you can get the latest version (currently
2.6.1pl4 (the four patches are separate, they are available from my
mail-server as gs261-patches-1-thru-4.tar.gz.uue)) without waiting for
the SLS to get it in there.  Relying on softlanding will make your
system inefficient; you certainly don't need drivers for every printer
there is in your gs binary.  You should learn how to build things
under unix yourself.  If you haven't the time, others do, and post
binary installations independently from the SLS folk to major linux
ftp sites.  (This is not an SLS bash; the SLS is most complete way to
bootstrap yourself into the Linux world, but it is not very good for
upgrading purposes).  One of these options may be needed if you don't
have X libraries; the SLS binaries usually include the X11 device and
therefore want the X libraries.

A minimal binary installation of ghostscript and several other
packages needed for printing the linux documentation (which is in the
works and comes in LaTeX - now in partial alpha release (!) and very
well done, if I may say) is available as:

sunsite.unc.edu:/pub/linux/apps/tex/texmin/texmin-0.1.tar.z

Note that this does not contain any postscript fonts.  (Nor do you
need them to print dvi (aka [La]TeX).)

The main Ghostscript documentation is contained in the file use.doc,
either in the source directory, or ...lib/ghostscript/doc/use.doc if
you haven't the sources.

To print ps, first determine your driver name with:

gs -help

which lists installed drivers.  If your favorite device is not listed,
you must compile gs yourself from the source distribution (do not
panic.  do follow the instructions in make.doc.  you will need 5 or 6
megs of space to build it in.)  Then type

gs -dNOPAUSE -sDEVICE=?????? -sOutputFile=/dev/???? file.ps

and your output should (hopefully) appear at your printer.  Those of
you with non-US paper sizes may wish to build gs yourself with the
right default, or you may use the -sPAPERSIZE=a4 option.

Ghostscript may be used to print at most of the resolutions your
printer supports; -r300, -r150, -r360x180 are examples of the option
used to control this.  Dot matrix printers in general and 24 pin Epson
compatibles in particular need to choose an appropriate resolution.
Note that versions 2.6 and greater of ghostscript have more drivers.

------------------------------

Subject: fonts
From: [e-mail]

(This section contains font information not specific to any
ghostscript driver (nor even specific to ghostscript).  Font
information pertaining to the X11 gs drivers (and thus to ghostview)
is included in the ghostview section under previewing)

All versions of ghostscript come with assorted public-domain fonts,
most of which were generated from bitmaps.  (meaning that they look
bad).  However, ghostscript can use any Type 1 or 3 fonts you may have
handy.  For example, the Adobe Type Manager (for any platform other
than mac (a conversion util for mac ps fonts is on simtel in the
windows3 directory)) comes with fonts you can use.  (get out some of
that old windoze software :) .  There may be some licensing
considerations, but if you're not a large company, no one will
notice.  Place the fonts (*.pc?) in .../lib/ghostscript/fonts/ and add
to your fontmap lines such as:

%(lines toward the end override earlier entries, so put these at the end)
%(this is taken from the ms-dos distribution of gs on simtel-20)
%(font names must be the same at in the actual font, but you may use
% aliasing in this file to use a different font, read the docs)

% Fontmap.ATM - an alternative Fontmap for Ghostscript,
% suitable for use with Adobe Type Manager fonts.

% Before you use ATM fonts with Ghostscript, please read carefully
% the license that accompanies the ATM fonts; Aladdin Enterprises and
% the Free Software Foundation take no responsibility for any possible
% violations of such licenses.

% ATM fonts.

/Courier                        (com_____.pfb)  ;
/Courier-Oblique                (coo_____.pfb)  ;
/Courier-Bold                   (cob_____.pfb)  ;
/Courier-BoldOblique            (cobo____.pfb)  ;

/Helvetica                      (hv______.pfb)  ;
/Helvetica-Oblique              (hvo_____.pfb)  ;
/Helvetica-Narrow               (phvrrn.gsf)    ;
/Helvetica-Bold                 (hvb_____.pfb)  ;
/Helvetica-BoldOblique          (hvbo____.pfb)  ;

/Symbol                         (sy______.pfb)  ;

/Times-Roman                    (tir_____.pfb)  ;
/Times-Italic                   (tii_____.pfb)  ;
/Times-Bold                     (tib_____.pfb)  ;
/Times-BoldItalic               (tibi____.pfb)  ;

TimesNewRomanPS and ArialMT fonts are interchangable with Times Roman
and Helvetica, so you can alias them this way if this is what you
have.

Adobe Type 1 fonts may be found on the net:

         ftp.cica.indiana.edu:/pub/pc/win3/fonts
         ftp.cica.indiana.edu:/pub/pc/win3/fonts/atm
         archive.umich.edu:/msdos/mswindows/fonts

I have not looked in these places, but this information is lifted
straight from the comp.fonts HOWTO (which you should read if fonts are a
thing for you.  I'll be glad to mail you a copy).

Conversion between various font types is tricky.  Ghostscript comes
with the tools needed to take a bitmap (hopefully large) and make a
scalable ps font.  Groff comes with the tools to allow use of tfm
(TeX) and pfb (Type 1) fonts in *roff documents.  X11R5 includes
several font utilities and Type 1 rendering code contributed by IBM.
I have used none of the above, but they all come with manpages, so
read them.  Other font conversion utilities are listed in the
comp.fonts HOWTO.  Also look into the package fontutils on
prep.ai.mit.edu:/pub/gnu

------------------------------

Subject: faxing
From: [e-mail]

If you have a fax modem, you can arrage to fax things out (and in),
including PostScript, dvi, ascii, etc...  Arranging for e-mail to fax
itself somewhere is also quite possible.

Fax modems correspond to one of two types: Classes 1 and 2.  Class one
modems are fairly common and have less of what goes on in a fax
supported in firmware (thus the software has to do more and do it
carefully.  supporting such timing-critical things under a preemptive
multitasking environment such as Linux is tricky at best).  The class
1 standard is EIA 578.  Class 2 modems tend to be more expensive and
comply with the standard EIA 592.  Mention of your fax modem's level
of support should be in its manuals.  Do not confuse class and group.
You would be hard pressed to find a non-group III fax modem these days
(but you probably could).

Most (all?) fax software which runs under linux must convert input in
whatever format into a Group III compatible image format for
transmission.  As usual, Ghostscript does the job.  The devices
dfaxhigh and dfaxlow generate faxable output at the two common fax
resolutions (200x200 and 200x100).  tiffg3 generates standard g3/tiff
encoded fax messages.  You must compile them in if they are not there
already; this is yet another reason to get the source.  Some fax
sofware on commercial platforms can use the display PostScript
renderer to render the fax image; this is not yet an option under
Linux.

The GNU program netfax supports Class 2 fax modems only.  It builds
nearly out of the box under linux.  I have patches for version 3.2.1.
It runs a server and accepts jobs in a queue from the local machine
and the net.  netfax is in use at MIT.  It accepts postscript, dvi,
ascii, and can be configured to work as an email gate.  Documentation
in the package is limited to compile notes and man pages for the
individual programs; actual installation tips are few and far
between.  But this is nothing new in linux :)

prep.ai.mit.edu:/pub/gnu/fax-3.2.1.tar.gz

FlexFax is available from:

sgi.com (192.48.153.1):/sgi/fax/?????.src.tar.Z

There is a beta containing a linux port; it built happily for me (but
I haven't tried it yet).  It is written in C++, and thus requires g++
to build.  It also links statically and seems to end up as a curiously
large binary.  FlexFax supports class 1 and 2 modems, uses ghostview
or a display postscript server for rendering, and generally is more
complete, or at least general than the somewhat MIT-specific netfax.
It also can run in place of getty and either receive a fax or spawn a
login as appropriate for the type of call.

mgetty+sendfax is a Linux and SCO-specific getty for faxmodems and a simple
sendfax program.  This package is on sunsite.

modgetty (which uses tcl, an interpreted language) can receive faxes.
Try harbor.ecu.purdue.edu:/pub/tcl

efax is in beta, it offers simple fax functionality for Class 2 fax
modems under Linux.

------------------------------

Subject: {n,t,g}roff, man pages:
From: [e-mail]

Man pages can sometimes be printed straight from the cat pages which
come with the SLS (in lieu of the normal nroff source pages) as though
they were a normal text file (which they mostly are).  However, many
printers do not like the VT escape codes for highlighting and what not
that find themselves imbedded in these.  A filter for this purpose
comes with the net-2 lpd package.  It is called lpf.

If you have the nroff source to the page (the finding of which I
highly recommend) you can say:

man -t foobar |lpr

and your man program will (hopefully) format the man page using groff
into PostScript, which will then be sent to your lpd and on to the
printer.  This form of man page output looks MUCH better than the
plain ASCII version.

If your man doesn't do this, you might try the perl version of man,
available near:

sunsite.unc.edu:/pub/Linux/system/Manual-Pagers/man-pl*

It is written entirely in perl, and is thus easily customizable (perl
being an interpreted language reminiscent of C and sh).

You can also find the nroff source file in the man directories (most
mans have an option to just spit out the filename) and do it yourself,
into any format supported by groff:

groff -mandoc -T{ascii,dvi,ps,X100,X75,latin8} foobar.1  \
                                        [|{dvips,lpr}]

------------------------------

Subject: mpage
From: [e-mail]

The package mpage is an ASCII postscript-ifyer and n-up reformatter
for postscript.  It is available as:

wuarchive.wustl.edu:/pub/mirrors/unix-c/postscript/mpage.tar-z

or thereabouts.  Note that wuarchive uses the -z suffix to mean .Z,
ie, compress, not GNUzip or freeze.

man -t foobar|mpage

will send a 2-up (depending on the environment variable MPAGE) version
of the man page to lpr and its postscript interpreter.  This saves
paper and speeds up printing.

------------------------------

Subject: a2ps
From: [e-mail]

A2ps will take ASCII and turn it into a nice postscript document with
headers and footers and page numbers, printed two pages on one (or
otherwise, if you say so).  A2ps does a very nice job at this.  It is
available at the same place mpage is.

------------------------------

Subject: enscript
From: [e-mail]

Enscript is a program which does basically the same thing as a2ps.  I
do not know where to get it.  It comes with most unixes.

A clone version of enscript is called nenscript, available as
/pub/Linux/system/Printing/nenscript-1.13++.bin.tar.z or
/pub/Linux/system/Printing/nenscript-1.13++.tar.z

on sunsite (?).

------------------------------

Subject: gslp
From: [e-mail]

Gslp is one of the uilities which comes with ghostscript 2.6.x and
purports to do the same ascii --> ps conversion as enscript and a2ps.
I have not used it, but the docs say that

gs -q -sDEVICE=????? -dNOPAUSE -- gslp.ps text.file [options]

should do the trick.  (gslp.ps is the actual program, which is written
in postscript.  Here it is run with the argument `text.file.'
Postscript is in many respects a programming language more than a mere
printer command language.)  Further documentation is in the file
gslp.ps.  There is a script file which does the above for you.

------------------------------

Subject: psutils
From: [e-mail]

Those of you who deal with large amounts of postscript may wish for
more utility programs than come with the SLS.  There are probably
millions of little programs which do things to your Postscript. A
representative package of them may be found as:

achilles.doc.ic.ac.uk:/tex/inter/psutils/*
or guardian.cs.psu.edu: /pub/src/psutil.tar.Z

These handle page selection, double-sided printing, booklet creation,
etc.  Most large ftp sites (eg, wuarchive.wustl.edu, ftp.uu.uunet)
will have many such packages to choose from.

------------------------------

Subject: tex/dvi
From: [e-mail]

[La]TeX is the text formatting package used by many in the academic
world and many in other places where people realize word perfect to be
annoying.  TeX works much like any other compiler -- source code is
run through the program `tex' to become a .dvi file (analogous to an
.o object file in C) which can be further manipulated to produce
printed output.  This manipulation of a dvi (device independant) file
usually takes a little bit of doing.  This `bit of doing' is well
worth it; TeX's output is of professional quality.

For those in the real world who cannot afford a dvi understanding
printer, it is usually best to convert the dvi into postscript that
you can pipe into ghostscript or lpr.  The SLS comes with a
functioning installation of both TeX and dvips.  Typing

dvips -f1 file.dvi |lpr

will do it.  Dvips responds to either command line arguments or a file
/usr/TeX/lib/tex/ps/config.ps (in the SLS, at least) in which you can
arrange to have dvips automatically send its output to lpr.  Thus

dvips file.dvi

will do everything that needs to be done.  (this is much better than
the 'manual'

dvips -f1 file.dvi |gs -q -rxxx -dNOPAUSE -sDEVICE=xxx \
                                        -sOutputFile=- - |/dev/????)

Note that some .dvi's may include PostScript graphics in the dvips
stage rather than the tex stage of the game; if they are not found,
you will get a hole instead of a picture.  This follows naturally from
the object file analogy above :) Usually, pre-made documentation in
this form has a makefile or script.  Lilo docs are an example of
this.

Dvips has several interesting options; for exemple:

dvips -r1 file.dvi

will print it out backwards.  We deskjet users love this one.

If all you are given is a file with a .tex ending, try either tex
file.tex or latex file.tex.  One of these is bound to work.  Then you
have a dvi.  (You may have to run it twice for indexing)

------------------------------

Subject: dvilj
From: [e-mail]

For LaserJet owners, there is a separate program which I beleive is in
the SLS that will take dvi and convert it directly into your printer's
language (PCL).  It is called dvilj2p or dvilj, and if not on tsx or
sunsite, is certainly available on ftp.uu.net.

(Description by Nils Rennebarth)

Its a nice driver, but a little out of fashion in the sense that
configuration (especially of font-paths) font-paths is not very
flexible and that it doesn't support virtual fonts (at least the
version 0.51 not).  The biggest advantage over the dvips/ghostscript
combo is that it uses downloadable fonts which:
  a) reduces data transmission to the printer drastically, which
     makes the printer usable even on a serial line.
  b) reduces printer-memory-requirements drastically. A standard
     Laserjet with 512k memory is able to print almost every
     TeX-document.

It has support for double side printing and most options you expect a
driver to have. It compiles cleanly and worked flawlessly on our
diverse hardware here.

There are other specific dvi drivers for different printers, as well.
I know of ones for the HP Deskjet and Epson.

------------------------------

Subject: texinfo
From: [e-mail]

This is the native documentation format of the GNU project.  Emacs can
be coerced into producing an info file out of this, and TeX can
produce nice printed docs from the same file.  It is a bit of a
stretch for both systems, but it works.  It is really just TeX source
which expects the macro file texinfo.tex to be installed on your
system.  The SLS ought to have the macro package.  Just do:

tex thefile.texi[nfo]

twice (for index generation purposes), and you end up with a plain
.dvi, to print or preview at your leisure.

In emacs, you can also do M-x texinfo-format-buffer to convert the
texinfo file into an info file viewable with emacs M-x info or an info
viewer of your choice.

------------------------------

Subject: hardware and drivers
From: [e-mail]

(This information is based on perusal of the source to Linux 0.99p9
and a letter from Michael K. Johnson.)

There are two ways the kernel driver may be used to run the parallel
ports.  One, the original, is the polling driver.  The other is the
interrupt driver.  In principle, the interrupt driver only deals with
the port when it gets an interrupt and should therefore be more
efficient.  In practice, people have found that it depends on the
machine.  It probably doesn't make too much difference in most
situations.

For the polling driver, you may adjust its polling frequency with the
program lptune without kernel twiddling. The actual driver is in
.../linux/kernel/chr_drv/lp.c.

To choose the interrupt driver rather than the polled, use the program
lptune (or lpcntl) to set it.  (Lptune (and lpcntl) is available on
sunsite, or from my mail-server.)  Just put the appropriate line in
/etc/rc.

You can also change it in the kernel if you really want to.

Seven is the usual "LPT1:" IRQ, 5 is "LPT2:" for an AT class machine.
Note that if your printer is on lp0, the interrupt driver will
probably not work.  The first parameter should already contain the
correct i/o base address.  Many bios boot screens have this
information if you need it.

------------------------------

Subject: Printer device names
From: [e-mail] (Brian McCauley)
Reply-to: [e-mail] (Grant Taylor)

On an XT bus system LPT1: becomes /dev/lp0 (major=6, minor=0),
on an AT LPT1: becomes /dev/lp1 (major=6, minor=1). To be more
precise:

Name Major Minor I/O address
 lp0   6     0       0x3bc
 lp1   6     1       0x378
 lp2   6     2       0x278

For a serial printer use the ttyS? (or ttys?, if you follow that
scheme) device, not the cua?  device. The ttyS? devices have major 4;
the cua? devices have major 5.

------------------------------

Subject: LPR                <<<<<<------------
From: [e-mail] (Brian McCauley)

This section used to the the lpd-HOWTO then it became part of the
linux-HOWTO and now it is part of printing.how-to. Who knows where it
will go next?
--
    \\   ( )   No Bullshit!   | Email: [e-mail]
 .  _\\__[oo       from       | Voice: [phone] (home)
.__/  \\ /\@  /~)  /~[   /\/[ |        [phone] (work)
.  l___\\    /~~) /~~[  /   [ |   Fax: [phone] (work)
 # ll  l\\  ~~~~ ~   ~ ~    ~ | Snail: 197 Harborne Lane, B29 6SS, UK
###LL  LL\\ (Brian McCauley)  |  ICBM: 52.5N 1.9W

------------------------------

Subject: What lpr and lpd do
From: [e-mail] (Grant Taylor)
Reply-to: [e-mail] (Brian McCauley)

Most unix systems use lpd (or the System V variant lp), the line
printer daemon, and friends to spool print jobs and run them through
all needed filters.  While line printers are certainly on their way
out, spooling print jobs and running them through filters is
definitely a convenient thing.  Thus lpr.  (subliminal note: PLEASE
find and read the lpr, lpd, printcap, lpc, lpq, and lprm man pages.
They are in the source dist of the net-2, if you haven't got them.)

Lpd and family are available in several places.  The new Linux net-2
package contains a working lpd.  I now use this one with my stock
0.99pl10 kernel after following the directions in Matt Welsh's net-2
HOWTO to get my NET-2 going (you only need loopback).  A binary package
which can be coaxed into working with the old net things (pre 0.99pl10
kernel) is in the SLS.  Note a subtle difference between the two
versions: the old one placed lock files and such in /var/spool, while
the net-2 version requires the directory /var/spool/lpd to exist for
it's lock file.

You will need to edit /etc/printcap to configure lpd to use your
printer. Setting up lpd to accept postscript and print it is not very
difficult; simply make a shell script fiter (qv).

Setting up lpr correctly is definitely worth the trouble if you are
going to do any printing at all.  Of course, if all you ever print is
the occasional man page or instruction book, then you can probably get
by without it (but even then it's nice).

Ghostscript 2.6.x also comes with a complicated script and utilities
to generate printcap entries and manage multiple types of queues for a
printer.  I have not tested these, but I assume they can be made to
work.  If you are in a large-scale environment with varied printing
needs, these are probably worth a look.

------------------------------

Subject: Where do the files go and what should their permissions be?
From: [e-mail] (Brian McCauley)

There is quite a bit of variation between the various releases
but I'll try where possible to offer solutions that are universally
applicable. Put /etc/lpd in your /etc/rc or /etc/rc.local (usually in
/etc/rc.local after you start syslogd (if you use syslogd)).

Set the group fields of the file permissons/ownership as follows:

-rwxr-s---   1 root     lp          37892 Nov 19 23:32 /etc/lpd
-rwx--s--x   1 root     lp          21508 Nov 19 23:32 /usr/bin/lpc
-rws--s--x   1 root     lp          17412 Nov 19 23:32 /usr/bin/lpq
-rwx--s--x   1 root     lp          17412 Nov 19 23:32 /usr/bin/lpr
-rwx--s--x   1 root     lp          17412 Nov 19 23:32 /usr/bin/lprm
-rwxr-xr-x   1 root     lp           2816 May 10 13:37 /usr/bin/lptest

...and for each of the spool directories listed in the sd fields of
/etc/printcap...

/var/spool/lpd:
total 5
drwxrwxr-x   2 root     lp           1024 May 18 23:00 .
drwxr-xr-x  11 root     root         1024 Feb 19 20:56 ..
-rw-rw-r--   1 root     lp              4 May 18 23:00 .seq
-rw-rw-r--   1 root     lp             18 May 18 23:00 lock
-rw-rw-r--   1 root     lp             25 May 18 23:00 status

Note these 3 files are created by lpr and lpd so if you've never
run these they will be absent and you should touch them into being.

In older versions the group was ``daemon'' not ``lp''. There's also
a named socket but lpd creates/deletes it as needed.

Some of these permissions may be overkill so don't be surprised if
your system works with different permissions. The fact that lpq needs
to be suid(root) is inelegant and may change in the future. People tell
me that lpr must be suid(root) but I've not seen evidence that this
is really the case as long as the file permissions on the spool queues
are right. Still as far as I know lpr is designed to be secure when
installed setuid(root).

Alternative approach: just make lpc, lpr and lprm setuid(root) then
you can almost forget the file permissions on the spool queues!

You're free to choose different directories for the executables on
your system (notably lpc is usually in /etc even though it has
commands that are useful to non-root).

The master lpd lock file is fixed at compile time to be
/var/spool/lpd/lpd.lock so you must have a /var/spool/lpd directory
even if you choose not to keep your spool queue there. In the older
binaries the lock file was /usr/spool/lpd.lock so this was not an
issue.

My advise is keep your primary spool queue in /usr/spool/lpd and make
/var a symlink to usr or keep it in /var/spool/lpd and make /usr/spool
a symlink to ../var/spool. This gives the greatest compatibility with
the pathnames that are compiled into the various distributed binaries.

The main configuration file is /etc/printcap although the old binaries
look for it in /usr/etc/.  The old binaries' README says hosts.equiv
and hosts.lpd files are looked for in /etc/inet but again the more
accepted convention is to locate them in /etc.

Since, IMHO, working out in which particular etc directory each file
should reside is tedious, I would advise having only one:

mv -i /usr/etc/inet/* /etc/inet/* /usr/etc/* /etc/
rmdir /usr/etc/inet /etc/inet /usr/etc
ln -s ../etc /usr/etc
ln -s . /etc/inet

[But make sure the files in /usr/etc and /etc/inet are not already
symlinks into /etc or you'll be in deep water.]

This makes your /etc huge but I think it's safe. All the latest stuff
has only /etc paths compiled into it so eventually we can get rid of
these links.

------------------------------

Subject: What are all these diretories for?
From: [e-mail] (Brian McCauley)

People like to put all things that aren't needed during bootstrap in
/usr so that / is as small as possible. They therefore move
unessential big files (usually the binaries) from /etc/ to /usr/etc/.

Unless you know what you are doing (or following the example of
someone that does) I'd advise keeping it all in /etc with /usr/etc and
/etc/inet being symlinks back to /etc. (In fact these links are
probably soon to be redundant as most recent binaires now look in
/etc/.)

The Linux/PRO approach is to have a real /usr/etc directory containing
all the non-essential stuff and make symlinks to each file therein
from its namesake in /etc. To spread out the configuration files for
the sake of tidiness there is a directory tree called /conf and many
files in /etc symlinks into this tree.

Also a few people like all their internet stuff in /etc/inet but as
recent binaries excpect to find this stuff in /etc they must create
links for each file from /etc into /etc/inet.

To avoid the problem of your /usr filesystem filling up as a result of
stuff in spool queues some poeple perfer to keep their spool queues
and such like in a separate filesystem called /var. If you don't have
a /var file system then you should make /var a symlink to usr. If you
_do_ have a /var filesystem to may want to make /usr/spool a symlink
to ../var/spool so that any programs expecting to find /usr/spool will
not be dissapointed.

------------------------------

Subject: lpd not working
From: [e-mail] (Brian McCauley)

If ``ps ax'' does not reveal a lpd then you daemon has died (or was
never started) - usually a sign that it couldn't create its lockfile
(or you tried to start it before all your filesystems were mounted).

If lpr works only for root then you've probably got a permission problem.

If you get ``jobs queued, but cannot start daemon'' or ``lpc: connect:
No such file or directory'' you are having trouble with the socket
connection to lpd. "start" in the context of this error really means
"wake". I don't really understand this but the problem was common in
early days with the old binaries but developments in the TCP/IP code
in the kernel fixed it for most people. Basically is stems from a
network porblem. If you're not on a network then try congifuring
youself as a single node network and the problem should vanish.

------------------------------

Subject: Where do I get a printcap for a xxxxx?
From: [e-mail] (Brian McCauley)

This question is essentially meaningless - see next question.

------------------------------

Subject: The semantics of /etc/printcap
From: [e-mail] (Brian McCauley)

Given the similarity in appearence and name between /etc/termcap and
/etc/printcap one could be forgiven for assuming that they contain
analogous infomation. This is not the case. Whereas termcap contains
informations about terminal *types* - (mostly escape seqences)
printcap contains information about *specific* printers (like the
directory that holds the spool queue and the device name of the
printer). The information about a printer model's escape sequences and
so on are held in the various filters which are _programs_ called by
lpd to drive the printer. /etc/printcap simply gives the locations of
these filters.  For details RTFM(printcap). [Alternatively the net-HOWTO
has a summary of some of the more important fields.]

One last point you should always specify ``suppress header'' ``:sh:''
unless you have *text* (_not_ postscript) printer and want banners. On
a text printer they are usually a waste of time and paper. On a
postscript printer they usually stop your printer working.  (See the
banners question for further expanation).

------------------------------

Subject: The syntax  of a /etc/printcap
From: [e-mail] (Brian McCauley)

Ideally RTFM(termcap) (yes, I said *termcap*) but since most people
don't have TFM(termcap) here are the essentials.

Lines starting # are comments (as you might have guessed).

For each printer usable from the lpr command on your system there is
one logical line in the file. For the sake of readability each logical
line may be spread over several physical lines by making the last
character on all but the last physical line a backslash.

Each logical line has the following format:

name1|name2|name3:string_capability=string:\
       :numeric_capability#number:boolean_capability:

The leading spaces and colon on the second line are for readability
only.

A printer can have as many names as you like but conventionally the
final name is used as a longhand description of the printer. (Still
people are free to say `lpr -P "Fred's grotty teletype"' if that's
the description you've given.)

The list of capabilities can be as long as needed and the order is not
significant. Each ``capability'' is denoted by a two character code.
(The name ``capability'' comes form the file format's termcap heritage
- parameter or attribute would be a more sensible terms.) [Note from
Ross Biro: capabilities with 3 character names don't work properly
which is why the serial port stuff in the old binaries failed.]
Capabilities having string value and have a = delimiter between the
capability name and the value while those having a numeric value use a
# (actually they can use either a # or an =). Boolean ``capablilties''
are true if they appear in the list and false if they do not.

Special characters in a string value can be expressed using `\'-escape
sequences as in C; in addition, `\E' stands for ESC.  `^' is also a
kind of escape character; `^' followed by CHAR stands for the
control-equivalent of CHAR.  Thus, `^a' stands for the character
control-a, just like `\001'. `\' and `^' themselves can be represented
as `\\' and `\^' respectively. `\:' for `:' seems to work but the
source code contains a waring that it can confuse the parser and
`\072' is a better idea.

Example:

lp|bam|BAM's Epson FX-80:lp=/dev/lp1:sd=/usr/spool/lp1:sh:mx#0:\
        :df=/usr/local/lib/magic-filter/lp.df:\
        :if=/usr/local/lib/magic-filter/lp.if:

Name is lp (this is the printer that lpr uses by default). It's also
known as bam or "BAM's Epson FX-80".

The printer is on /dev/lp1 (aka AT-bus LPT1:). I don't want a burst
page. I don't want a file length limit. Files queued by `lpr -d' are
passed through /usr/local/lib/magic-filter/lp.df and those queued by
`lpr' through /usr/local/lib/magic-filter/lp.lf.

See also the next section.

------------------------------

Subject: An /etc/printcap gotcha.
From: [e-mail] (Brian McCauley)

Two /etc/printcap files can look identical and yet one works and the
other doesn't.

See if ``lpc stat'' reports a printer called ` :'. The last
character on a continued line must be a \. If there are whitespace
characters after the \ then it doesn't register the next line as a
continuation.

------------------------------

Subject: The minimum /etc/printcap?
From: [e-mail] (Brian McCauley)

This is a silly question but it _is_ frequently asked. The answer is
``lp:sh'' (that's 6 bytes including the required linefeed character on
the end). To use this /etc/printcap you must make /dev/lp a symlink to
your printer and create your spool queue directory as /usr/spool/lpd.
(You might think that if you wanted banner pages you could loose the
``:sh'' but the termcap syntax requires at least one capability per
entry).

------------------------------

Subject: How do I prevent the `staircase effect'?
From: [e-mail] (Brian McCauley)

Unix terminates each line of a file with a linefeed but not a
carriage return so taken literally a Unix text file printed on an
ASCII device will start each line below the end of the previous line.
Some printers can be set to treat "linefeed" as "carriage return,
linefeed", others can't. If yours can then do simply do that. If the
printer cannot be fixed create a shell script filter that reads:

#!/bin/sh
if [ $1 = -c ]; then
  cat
else
  sed -e s/$/^M/
fi
# the ``echo -ne'' assumes that /bin/sh is realy bash
echo -ne \\f

where ^M is a carriage return character not a ^ followed by a M.  To
type ^M in emacs use the sequence C-q C-m and in vi use C-v C-m.
Conventionally this script is called /usr/lib/lpf. The test of $1
allows the isertion of carriage returns to be switched off by the `-l'
switch on lpr.

Alternatively your printer may have an escape sequence that will set
the way it handles linefeed characters. A simple filter that uses an
`echo -ne' command to send this sequence may be appropriate.

#!/bin/sh
# Filter for HP printers to treat LF as CRLF
# the ``echo -ne'' assumes that /bin/sh is realy bash
echo -ne \\033\&k2G
cat
echo -ne \\f

------------------------------

Subject: Resetting the printer after each printout

Either make your filters do it or define the `tr' ``capability'' in
/etc/printcap to be your printer's font reset command. For details of
the format of this string see the question on the format of
printcap. This may not work if a printout crashes in the middle of an
escape sequence - putting a lot of ^@ on the front may help but this
probably won't be enough it you were printing raster graphics when the
filter died.

------------------------------

Subject: Preventing formfeed after each printout
From: [e-mail] (Brian McCauley)

If you don't have an `if' specified in /etc/printcap then lpd will
automatically put a formfeed at the end of each file. If you're using
a filter then it's up to the filter to decide if it wants to put a
formfeed. To disable formfeed completely if you don't have an `if' put
:ff=: in your /etc/printcap.  But please note this suppresses the
formfeed that would usually be printed if a filter dies. If you want
formfeeds after text printouts but not on printouts printed with `lpr
-l' then create the following `if' filter:

#!/bin/sh
# the ``echo -ne'' assumes that /bin/sh is realy bash
cat
if [ "$1" != -c ]; then
  echo -ne \\f
fi

If you want a formfeed after `lpr -l' to be optional you can misuse the
`-i' switch to suppress the formfeed with the following trick (after
all `lpr -i -l' would usually not be implemented).

#!/bin/sh
cat
# use lpr -i -l to print raw without trailing formfeed
if [ "$1" != -c -o "$4" = -i0 ]; then
  # the ``echo -ne'' assumes that /bin/sh is realy bash
  echo -ne \\f
fi

------------------------------

Subject: Printing to a serial port
From: [e-mail] (Andrew Tefft)
Reply-to: [e-mail] (Brian McCauley)

The first section applies to both net-2 and pre-net-2 implementations
of lpd; then there is a section applicable only to the older
implementation; the rest applies to net-2 only.

First, there is a kernel patch you need to make. This applies to
pre-net-2 lpd as well (actually, the use of this should probably
just be commented out of the lpd source code). The change is for
at least patchlevel 10, but probably all others as well.

in /usr/src/linux/kernel/chr_drv/tty_ioctl.c, around line 390,
you will see:

                case TIOCEXCL:
                        return -EINVAL; /* not implemented */

Change this to return 0, recompile the kernel, and you'll be all set.

The following applies only to the "old" lpd, before net-2.

You cannot set any serial port settings (including baud rate) in your
printcap. Leave them all out and use stty to set up the port (if necessary)
in your rc.local, and you should be all set (to find the correct
stty settings, start with the default and twiddle until you can correctly
cat a file out the port).

The rest of this applies only to net-2 lpd although it may be
helpful to read. I put more detail into this because it's what I use.

------------------------------

Subject: Serial settings in printcap
From: [e-mail] (Andrew Tefft)
Reply-to: [e-mail] (Brian McCauley)

There are two sets of flags which you will need to set, plus
the baud rate (Note: the "fc" flag setting seems to override the br#
capability, so be sure to set that correctly as well as the br#!).

Each of the flags can have bits set and cleared. Clearing is done
first, so specify the clear flags (fc# and xc#) before the set flags
(fs and xs).

   Setting the br# capability is self-explanatory. Example: br#9600

   It is very easy to translate from 'stty' settings to printcap flag
   settings. If you need to, see the man page for stty now.

   Use stty to set up the printer port so that you can cat a file
   to it and have it print correctly. Here's what "stty -a" looks like
   for my printer port:

dina:/usr/users/andy/work/lpd/lpd# stty -a < /dev/ttyS2
speed 9600 baud; rows 0; columns 0; line = 0;
intr = ^C; quit = ^\; erase = ^?; kill = ^U; eof = ^D; eol = <undef>;
eol2 = <undef>; start = ^Q; stop = ^S; susp = ^Z; rprnt = ^R; werase = ^W;
lnext = ^V; min = 1; time = 0;
-parenb -parodd cs8 hupcl -cstopb cread -clocal -crtscts
-ignbrk -brkint -ignpar -parmrk -inpck -istrip -inlcr -igncr -icrnl ixon -ixoff
-iuclc -ixany -imaxbel
-opost -olcuc -ocrnl -onlcr -onocr -onlret -ofill -ofdel nl0 cr0 tab0 bs0 vt0
ff0
-isig -icanon -iexten -echo -echoe -echok -echonl -noflsh -xcase -tostop
-echoprt -echoctl -echoke

   The only changes between this and the way the port is initialized
   at bootup are -clocal, -crtscts, and ixon. Your port may well be
   different depending on how your printer does flow control.

   Once you have your stty settings right, so that cat file > /dev/ttyS2
   (in my case) sends the file to the printer, look at the
   file /usr/src/linux/include/linux/termios.h. This contains a lot of
   #defines and a few structs (You may wish to cat this file to the
   printer (you do have that working, right?) and use it as scratch paper
   -- I did!). Go to the section that starts out

/* c_cflag bit meaning */
#define CBAUD   0000017

   This section lists the meaning of the "fc#" and "fs#" bits.
   You will notice that the names there (after the baud rates) match
   up with one of the lines of stty output. Didn't I say this was going
   to be easy?

   Note which of those settings are preceded with a - in your stty
   output. Sum up all those numbers (they are octal). This represents the
   bits you want to *clear*, so the result is your fc# capability.
   Of course, remember that you will be setting bits directly after
   you clear, so you can just use fc#0177777 (I do).

   Now do the same for those settings (listed in this section) which
   do not have a - before them in your stty output. In my example
   the important ones are CS8 (0000060), HUPCL (0002000), and CREAD (0000200).
   Also note the flags for your baud rate -- mine is 0000015. Add those
   all up, and in my example you get 0002275. This goes in your fs#
   capability (fs#02275 works fine in my example).

   Do the same with set & clear for the next section of the include file,
   "c_lflag bits". In my case I didn't have to set anything, so I just
   use "xc#0157777" and "xs#0".

   Once your printcap is set up, try it out. If things don't work,
   see the next section.

------------------------------

Subject: cat works to the serial port, but not lpd (1)
From: [e-mail] (Andrew Tefft)
Reply-to: [e-mail] (Brian McCauley)

Generally getting lpd up and running is explained elsewhere, but if
you are having trouble with serial port settings you can prevent lpd
from trying to configure your port by treating you printer as one that
does not present a normal device interface (qv).

   1. Set your printer (in your printcap) to /dev/null1. Create /dev/null1
      using mknod /dev/null1 c 1 3 (because you don't want /dev/null to be
      opened exclusively).

      remove the baud rate and flags settings from your printcap.

   2. Create a script such as this:

#!/bin/sh
echo if: $* >> /var/spool/lpd/results
# /dev/lp is linked to /dev/ttyS2 which has the printer
exec your-old-input-filter $* > /dev/lp

   ...or if you didn't have an old if installed...

#!/bin/sh
echo if: $* >> /var/spool/lpd/results
cat > /dev/lp
# the ``echo -ne'' assumes that /bin/sh is realy bash
echo -en \\f > /dev/lp

      Make sure it's world-executable and world-readable. I call my
      script /var/spool/lpd/if, to make things easy. Try out
      your script (/var/spool/lpd/if < somefile) and see if it prints.

   3. Set the if= capability in your printcap to call this script,
      e.g. if=/var/spool/lpd/if

   4. Use stty to correctly set your port settings. Try to print now.
      You should be able to tell if things are being spooled, and
      things *should* be printed, if your manual testing of the if
      script works. But this is a kludge, so the idea is not to use the
      if script.

------------------------------

Subject: cat works to the serial port, but not lpd (2)
From: [e-mail] (Andrew Tefft)
Reply-to: [e-mail] (Brian McCauley)

Assuming the above method using the "if" filter works and that
you belive that you have specified what you think are the correct
flags and baud rate in printcap; check ``stty -a < /dev/ttyS2'' (or
whatever your printer port is).  If the settings are not correct,
check your flags against your printout from termios.h. If the settings
are *way* not correct, you may need to fix lpd itself. The patch
follows, and you can probably see why it's needed :-) It gets applied
in the lpd/ subdirectory of the lpd source.

(the patch is coming in just a sec)

When I was setting mine up, I followed a sequence like this:

lprm (whatever -- make sure queue is empty and lpd is running)
stty (correct settings) < /dev/ttyS2
lpr (something small)
stty -a < /dev/ttyS2   (often had to ctrl-c out of this one)
(twiddle with flags)
lprm (whatever -- make sure queue is empty again...)

Here's the patch (I think its reversed so apply it with -R):
-------------------------------Cut Here-------------------------------------
*** printjob.c  Thu Jul  8 20:56:59 1993
--- printjob.c~ Sat Feb 27 09:07:01 1993
***************
*** 1271,1277 ****
        }
  #ifdef LINUX
        ttybuf.c_cflag &= ~FC;          /* not quite right! */
!       ttybuf.c_cflag |= FS;           /* not quite right! */
  #else
        ttybuf.sg_flags &= ~FC;
        ttybuf.sg_flags |= FS;
--- 1271,1277 ----
        }
  #ifdef LINUX
        ttybuf.c_cflag &= ~FC;          /* not quite right! */
!       ttybuf.c_cflag |= ~FS;          /* not quite right! */
  #else
        ttybuf.sg_flags &= ~FC;
        ttybuf.sg_flags |= FS;
-------------------------------Cut Here-------------------------------------

------------------------------

Subject: Printers that aren't simple character devices
From: [e-mail] (Brian McCauley)

[Firstly I'll explain the subject.] The most common example is a
printer that is connected via a network in some strange way.  For
example consider a printer conneccted to a host running some esoteric
OS with which you can only communicate via E-mail. To use this printer
without going through lpr you would say something like:

  | uuencode tmp | mail printer-request@eosteric

in place of:

  > /dev/lp

To use such a printer through lpr the ``lp'' capability of the print
queue should be directed to a /dev/null type device (e.g. ``mknod
/dev/null1 c 1 3'') but not /dev/null itself as lpd opens the device
exclusively. _Each_ filter must must explicitly uuencode and mail its
output. In more complex cases if you already have an ``if'' or ``of''
filter for a strangely connected printer then other filters can pass
their output to/through this filter to avoid duplication of effort.
In this case the ``if'' filter should usually be called with the
``-c'' switch to minimise the further manipulations if performs.

------------------------------

Subject: Getting burst/banner pages
From: [e-mail] (Brian McCauley)

For a simple text printer (in particular not postscript) and a
simple text banner simply take :sh: out of the printcap record. If you
want to prevent the banner comming out in whatever font was last used
on the printer then define the `tr' ``capability'' to be your
printer's font reset command.

If you want a fancy customised banner (or have a postscript printer)
leave :sh: in the printcap and make each of your filters print the
banner. All the information to put on the banner is included in the
filter's positional parameters.  RTFM(printcap) for details. [ If
you're using <[e-mail]>'s magic-filter package then
call the code to print the banners from the config script. ]

------------------------------

Subject: Printing text on a PS printer
From: [e-mail] (Brian McCauley)

You need a filter based on a program that converts ascii to
postscript. The most well known of these is ``enscript'' but it's also
the hardest to find. Others include ``a2ps'' and ``mpage'' (for
details of these see elsewhere in this document).

------------------------------

Subject: Trunation of graphics files
From: [e-mail] (Brian McCauley)

This is usually because you've got a limit set on the maximum size
file that can sit in the spool queue. Put `mx#0' in your printcap.

------------------------------

Subject: Why doesn't `lpr -i' work?
From: [e-mail] (Brian McCauley)

To get lpr -i to work you need a filter istalled as `if' that
implements it.  The -i switch is simply passed on by lpd to the filter.
There is such a filter usually called /usr/lib/lpf that can be obtained from
bsd source archives in the directory bsd-source/lpr/filters.  It's also part
of Fred's port but he put it in /usr/bin/.  If you whant to use this program
but still want your filter to do some printer specific initialisation then
write script thus:

#!/bin/sh
# My initialisation stuff goes here
exec /usr/lib/lpf $*

------------------------------

Subject: Why doesn't `lpr -p' work?
From: [e-mail] (Brian McCauley)

Because it's broken. lpd always thinks that the printer is 0
characters wide regardless of what printcap says. Previous versions of
the lpd-HOWTO contained a patch but it has now been applied to the
latest distribution of the net-2 stuff.

One other thing: lpd calls pr by full pathname so if you keep pr
somewhere different from /usr/bin/pr you will need a symlink.
(Where lpd expects to find pr may vary from version to version).

------------------------------

Subject: lprm says `too many users'
From: [e-mail] (Brian McCauley)

You have an anchient libc.so with a broken scandir().

------------------------------

Subject: lpc complains that it hasn't stopped a daemon
From: [e-mail] (Brian McCauley)

One lpd process runs all the time and it spawns children to
handle each printer as needed. Sometimes these children leave lock
files lying about. This is harless.

------------------------------

Subject: lpr over a network
From: [e-mail] (Brian McCauley)

To print on the printer listed as foor "foo" in the printcap on machine
"bar.baz.net" from the machine "mine.baz.net" you put an entry like
this in your /etc/printcap (on mine.baz.net):

foo:lp=:rm=bar.baz.net:rp=foo:sd=/usr/lpd/spool/foo:

and, of course, create the spool directory /usr/lpd/spool/foo.

There's no point specifying filters and the like in
mine.baz.net:/etc/printcap as it's the ones in
bar.baz.net:/etc/printcap that will get used.

On the machine bar.baz.foo, you need to put "mine.baz.net" on a line
by itself in either /etc/hosts.equiv or /etc/hosts.lpd; note that
putting it in /etc/hosts.equiv will allow for unauthenticated logins as
well as printing.  /etc/hosts.lpd is printing only.

------------------------------

Subject: Writing lpd filters
From: [e-mail] (Brian McCauley)

In normal Unix terminology, filters are just programs (so they
must have execute permission) that read a stream from their standard
input and write to their standard output.

lpd filters are filters in the sense that thay read STDIN and write to
STDOUT, but are unusual in that they may assume that their standard
input is a file and perform lseek() operations on it.

All lpd filters have a common command line syntax (or more often
simply ignore command line parameters). For details of the command
line parameters RTFM(printcap).

It's easier to debug filters if you test them in an immediate shell
before you install them. (If your filter makes use of its command line
arguments you'll have to specify them too).

my-new-filter <file >/dev/lp1

If you want to write a shell script filter it must have a #!/bin/sh
header.  Here is the generic form of a filter to accept postscript.

#!/bin/sh
/path.../gs -q -dSAFER -dNOPAUSE -r??? -sDevice=?????? -sOutputFile=- -

and place its full pathname as one of the filter (usually `if' but
_not_ 'of'!) parameter in the printcap for your printer.  I suggest
putting such scripts in a subdirectory off /usr/lib (sat
/usr/lib/lpd/). (-dSAFER attempts to protect against ps interpreter
security holes, -q and -dNOPAUSE make it churn right along quietly,
and Device is the appropriate driver for your printer).

Here is an Epson FX-80 dvi filter using ghostscript:

#!/bin/sh
/usr/TeX/bin/dvips -f | \
/usr/bin/gs -q -dSAFER -sDEVICE=eps9high -r120x216 -dNOPAUSE -sOutputFile=- -

More tools useful for makeing filters are described elsewhere in this
document.

------------------------------

Subject: A filter tests OK but fails when lpd uses it
From: [e-mail] (Brian McCauley)

You may need a #!/bin/sh header. You may also need to set PATH
within the script since the daemon's PATH may not have everything you
need.

------------------------------

Subject: `of' filters.
From: [e-mail] (Brian McCauley)

Never use these. (Well strictly speaking there are circumstances but
you're unlikey to meet them).

------------------------------

Subject: Getting filters for given printers
From: [e-mail] (Brian McCauley)

>From BSD source archives (I think). [Editors note: I need a
better answer here ].

If you already have a program to print say DVI on your printer by some
mechanism then making it into a filter is usually a matter of writting
trivial shell script - see this HOWTO. If the program you are using
insists on reading a names file as input see the next question. Text
mode filters are trivial too (see this HOWTO) unless you want lpr to
have a choice of fonts in which case they are slightly harder than
trivial.  You will probably want to insert and `echo -ne' command at
the beginning and end of your filter to set up the font etc to your
liking.

------------------------------

Subject: Filters from programs that won't read STDIN
From: [e-mail] (Brian McCauley)

For example dvilj2p insists on a named file as its input (and what's
more expects one with a .dvi suffix) so do this:

#!/bin/sh
ln -s /proc/self/fd/0 /tmp/$$.dvi
dvilj2p /tmp/$$
rm /tmp/$$.dvi

Note: If it wasn't for the fact that dvilj2p adds a .dvi suffix you
wouldn't need the temporary symlink and could just specify
/proc/self/fd/0 directly. People who use this trick often usually keep
permanant symlinks to /proc/self/fd/* in /dev. If you don't have
procfs you'll need to create a temporary file.

------------------------------

Subject: Having many filters
From: [e-mail] (Brian McCauley)

Historically the lpr command was created to support a finite set of
possible file types. You can, in fact, use any of the filters for any
reason. If you're never going to use Benson Varian raster files you
could use the -v switch for GIF files. You could even use -d for low
res and -v for high res. Remember that if you create a filter for a
file format that takes a long time to process then your printer may
sit idle between print jobs even when there are things in the queue.

If you are on a network remember that the filter setups go on the
print server.  One way to avoid running out of filter options is to
define several logical printers in /etc/printcap that all point to the
same physical one and put each filter in the ``if'' field of a
different printcap entry. This has the advantage that you can set the
PRINTER enviroment variable to choose your filter rather than having
to specify it on the command line each time. One small problem with
this is that you have no control over the order in which files from
separate queues are printed.

Another way to avoid running out of possible types is to use magic
filters.

------------------------------

Subject: What are magic filters?
From: [e-mail] (Brian McCauley)

Magic filters deduce their input files' types from `magic
numbers' (distictive byte patterns at particular offsets).  Magic
filters are usually perl scripts, shell scripts or C programs that
simply identify the file type then call the appropriate non-magic
filter. Blatent plug :-) I <[e-mail]> have a generic
magic filter bash script that selects the right filter to use based on
the output of the `file' command. With my magic ``dvi'' filter (and
3 associated non-magic filters) I can do things like:

lpr -d file1.dvi file2.div.Z file3.ps file4.texinfo.gz

(BTW confguring lpr to handle texinfo files getting a bit silly - see
previous question).

This is now at:
tsx-11.mit.edu:pub/linux/sources/usr.bin/magic-filter-0.3.tar.gz
(Although the release number will possibly change in future).

Magic filters should never specified as `of' as the output filter only
gets called once if a number of files are printed without a gap.

IMHO magic filters as `if' can also cause problems as they may prevent
you, say, _listing_ a PostScript or nroff file. (Other people disagree
as, if you are making your printer available on a network, it can be a
pain having people accidently listing ps files.)
The following is an example of a magic shell script which should take
either postscript or text and deal with it:

------------------------------

Subject: Magic filter examples
From: [e-mail] (Grant Taylor)
Reply-to: [e-mail] (Brian McCauley)

The following is an example of a magic shell script which should take
either postscript or text and deal with it:

#!/bin/sh
# This is based on a script I received from Scott Doty and which was
# written by Keith Walker.  Keith's script made use of the fact that
# lpd passes options to if:
#
#  <if> -w<width> -l<length> -i<indent> -n <user> -h <host> <accountingfile>
#
# to print text out well at any size.  This one does not.  These options
# are also handy if you want to do your own snazzy header page, much
# like NeWSPrint from Sun does (although running PostScript through
# the display server to get it interpreted is a bit much :)
#
# This is untested, but bits of it came from working programs, so it
# should work.
#
# gs will reset the printer anyway, so the this text setup doesn't matter
# setup should include the escape code for \n conversion, if applicable
#
printf "<printer setup for text printing (escape codes, etc)>"

read first_line
first_two_chars=`expr $first_line : '\(..\)'`

if [ "$first_two_chars" = "%!" ]; then # it's postscript

        /usr/bin/gs -dSAFER -dNOPAUSE -q -sDEVICE=??????? -sOutputFile=- -

else # it's plain text

        echo -n $first_line
        cat
        printf "\014"

fi

Note that for the paranoid, shell scripts run as someone other that
the user are a security hole, but they are illustrative.

Even more snazzy things can be done.  The following was the gzip'd,
uuencoded, C source to a full-featured program which automagically
recognizes text, dvi, postscript, and direct printer driving (aka
binary) and does the appropriate thing.  You will have to modify it to
reflect your printer rather than a LaserJet.  In particular, the
driver used for gs, and the use of dvips instead of dvilj2p would need
changing.  This approach also has less overhead as there is not
another incarnation of bash interpreting it.  This was sent to me by
Hal Brooks; he wins the award for best print spooler setup from those
who responded to my survey. :) Note that this program is also
somewhat in need of a real test, nor does it make any use of the
arguments passed to `if', but it works for him.

<This was a tad long; it is available from Grant Taylor's mailserver>

------------------------------

Subject: PREVIEWING          <<<<<<------------
From: [e-mail]

These sections describe various ways to preview things under Linux -
that is, how to view them in a way approximating their final form
without printing them out.

------------------------------

Subject: ghostview
From: [e-mail]

Ghostview, a companion program for gs, previews postscript on an X
display.  It also lets you select individual or ranges of pages from a
postscript document to print using lpr.  The new version, 1.5, has
fixed a few glitches which never bothered me but may make a difference
to you.  It also calls gs with the -dSAFER option and has a few more
resource and command-line options relative to 1.4.1.  A binary
installation of this is probably in the SLS.  The real installation is
from source

prep.ai.mit.edu:/pub/gnu/ghostview-xxx.tar.z

and is a clean build.  Ghostview requires gs to work.  The new version
of gs, 2.6.x, will use X display fonts in an effort to improve
legibility at low resolution (a previous failing of this pair relative
to commercial display-postscript based systems such as Sun's
pageview).  This works very well for me at least, at the expense of
exact character positioning (x fonts have different widths).  If you
prefer exact character representation (within the limits of your
display), you might wish to let gs render some Type 1 fonts you
install instead of using platform fonts (or the awful fonts gs comes
with.  To do this while in ghostview (or in any situation involving
the X11 driver), place:

ghostscript.useExternalFonts: false

in your .Xdefaults file, and the platform fonts will not be used.

This is part of a message posted to gnu.ghostscript.bug by Tim
Theisen <[e-mail]>:

(note that the usual Linux X-server, XFree-1.3, does contain a Type 1
rasterizer which I beleive was contributed by IBM.  I do not know if
XS3, X8514, or Xega include this.  Look and see :)

X Resources:

    Ghostscript now uses the X Toolkit to pick up X Resources.  Now
    ghostscript uses the standard X rules that allow more specific
    resources to override less specific ones giving users the full
    power of X resources to control the X11 driver.  It also allows
    system administrators to establish an application defaults file
    with resources specific to their ghostscript installation.

    The customization choices mentioned in make.doc have been moved into
    X resources and are now configured at run time rather than compile
    time.  Sorry, this section of make.doc did not get revised for the
    2.6.1 release.

    If useBackingPixmap is set, ghostscript will attempt to allocate
    a backing pixmap.  If one cannot be allocated, ghostscript will
    issue a warning and ask for backing store instead.  (Since there
    were insufficient resources for a backing pixmap, the X server
    may not provide backing store either.)


Color Handling:

    Color Handling was totally revamped for gs 2.6.

    Ghostscript first checks for a suitable standard colormap.  If you
    have static colormap in your X server, it would be best to store a
    standard colormap property on the root window describing the color
    layout.  Ghostscript will then be able to take full advantage of the
    device.  If you have a standard colormap installed, ghostscript will
    start slightly faster since it does not have to allocate colors for
    a cube or ramp.

    If no standard colormap is available, ghostscript will allocate an
    RGB cube or gray ramp.  Ghostscript tries for a 5x5x5 cube on a color
    device, and a 128 gray ramp on grayscale devices.  It will never ask
    for more than 1/2 of the colors for a RGB cube or gray ramp.  It also
    takes into account the number of significant bits per pixel.  (i.e.
    It won't ask for 128 gray levels if you only have 16 available.)

    Ghostscript will attempt to allocate colors that are off the color
    cube/ramp as the picture is being rendered.  Ghostscript will keep
    track of 256 dynamic colors.  After all these are allocated,
    ghostscript asks the X server directly.

    The foreground and background color can be set explicitly.  This
    is important for the visually impaired and when using the ghostview
    widget.

    Color Resources:

    palette(Palette): Default value: "Color".  Other allowable settings:
    "Grayscale", "Monochrome".  The palette resource is used to restrict
    the palette used for display.  One can set palette to "Grayscale" or
    "Monochrome" to see how a file would be rendered in grayscale or
    monochrome on a color display.  I use it to avoid dithering of gray-
    scale figures on a color display with 4-bit DACs.

    maxGrayRamp(MaxGrayRamp): Default value: 128.  Maximum number of
    gray levels that ghostscript will attempt to allocate.  (It won't
    try for more than this on an 8-bit pseudo color display even if you
    set it higher.)  Set this lower if you want a smaller ramp and
    would prefer ghostscript to use dynamic colors.

    maxRGBCube(MaxRGBCube): Default value: 5.  Maximum number of
    colors levels that ghostscript will attempt to allocate.  (It won't
    try for more than this on an 8-bit pseudo color display even if you
    set it higher.)  Set this lower if you want a smaller ramp and
    would prefer ghostscript to use dynamic colors.

    I believe these values to be a good compromise between dynamic
    allocation and fall back onto a fairly good color cube for dithering.

    You can use the foreground and background colors to accomplish
    "reverse video".  However, if you have a grayscale device, it may be
    better to reverse the gray ramp using the following PostScript code
    fragment:

    [{1 exch sub} /exec load currenttransfer /exec load] cvx settransfer


Platform Fonts:

    The X11 driver now support native X11 fonts.  If you have installed
    the HP XLFD font extensions into your font or X server.  Ghostscript
    will also be able to use platform fonts at rotations of 90 degrees,
    with mirroring, and anamorphic scaling.

    The X11 driver does most if its work silently.  You can get it to
    report when it is using an X11 font by setting the logExternalFonts
    boolean in your X resources.

    The X11 driver is setup to use the standard fonts distributed with
    X11R5.  We purchased the Adobe Type Manager and the Adobe Plus Pack.
    These font packages give all the fonts normally found in the Apple
    LaserWriter Plus.  The X11 driver is setup to handle these fonts as
    well.  (They are a superset of the bitmap fonts distributed with X11.)

    You may set the regularFonts, symbolFonts, or dinbatFonts resources
    if you have different fonts available.  Each font name must have 7
    dashes or it will be ignored.  Minimize the use of wildcards to
    promote faster matching.  (I once encountered an X server that took
    many seconds to do a font lookup when wildcards were carelessly used.)

    There is a different list of fonts for each common encoding.  Regular
    fonts may be accessed in standard or ISO Latin 1 encoding.  The bdf
    files that are distributed with X11 are in the ISO Latin 1 encoding.
    This leaves out the ligatures.  Luckily, the ligatures are present in
    the bdf files, but are not given an encoding, essentially commenting
    them out.  You can use the fixfont program from the xproof
    distribution (ftp.cs.wisc.edu:/pub/X/xproof.tar.Z, or
    ftp.x.org:/contrib/xproof.tar.Z) to reencode the bdf files
    and build X11 fonts that contain the ligatures (i.e standard encoding).

    If you have the Type1 fonts mentioned above, and you installed the
    Type1 rasterizer into you font or X server, you can use the appended
    fonts.scale to name your fonts so that ghostscript can find them.

    Font resources:

    useExternalFonts(UseExternalFonts): Default value: true.
    This resource controls whether X11 fonts will be used.

    useScalableFonts(UseScalableFonts): Default value: true.
    This resource controls whether scalable fonts will be used.
    If you have an outline scaler in your X server, you should have
    this on.  If you have an X terminal, you may get slightly better
    performance with this on.  If you have to use the X11 bitmap scaler,
    turn this off.  Fonts scaled by the bitmap scaler look worse than
    the default ghostscript fonts.

    logExternalFonts(LogExternalFonts): Default value: false.
    Controls whether to report when X11 fonts are being used.


The following fonts.scale makes all of the fonts of the Adobe Type Manager
and Adobe Plus pack available in standard and ISO Latin 1 encoding.
(We were able to purchase the above two packages at an educational discount
 price of $150.)

70
agw_____.pfb -Adobe-ITC Avant Garde Gothic-Book-r-normal--0-0-0-0-p-0-iso8859-1
agwo____.pfb -Adobe-ITC Avant Garde Gothic-Book-o-normal--0-0-0-0-p-0-iso8859-1
agd_____.pfb -Adobe-ITC Avant Garde Gothic-Demi-r-normal--0-0-0-0-p-0-iso8859-1
agdo____.pfb -Adobe-ITC Avant Garde Gothic-Demi-o-normal--0-0-0-0-p-0-iso8859-1
bkl_____.pfb -Adobe-ITC Bookman-Light-r-normal--0-0-0-0-p-0-iso8859-1
bkli____.pfb -Adobe-ITC Bookman-Light-i-normal--0-0-0-0-p-0-iso8859-1
bkd_____.pfb -Adobe-ITC Bookman-Demi-r-normal--0-0-0-0-p-0-iso8859-1
bkdi____.pfb -Adobe-ITC Bookman-Demi-i-normal--0-0-0-0-p-0-iso8859-1
com_____.pfb -Adobe-Courier-Medium-r-normal--0-0-0-0-m-0-iso8859-1
coo_____.pfb -Adobe-Courier-Medium-o-normal--0-0-0-0-m-0-iso8859-1
cob_____.pfb -Adobe-Courier-Bold-r-normal--0-0-0-0-m-0-iso8859-1
cobo____.pfb -Adobe-Courier-Bold-o-normal--0-0-0-0-m-0-iso8859-1
hv______.pfb -Adobe-Helvetica-Medium-r-normal--0-0-0-0-p-0-iso8859-1
hvo_____.pfb -Adobe-Helvetica-Medium-o-normal--0-0-0-0-p-0-iso8859-1
hvb_____.pfb -Adobe-Helvetica-Bold-r-normal--0-0-0-0-p-0-iso8859-1
hvbo____.pfb -Adobe-Helvetica-Bold-o-normal--0-0-0-0-p-0-iso8859-1
hvn_____.pfb -Adobe-Helvetica-Medium-r-Narrow--0-0-0-0-p-0-iso8859-1
hvno____.pfb -Adobe-Helvetica-Medium-o-Narrow--0-0-0-0-p-0-iso8859-1
hvnb____.pfb -Adobe-Helvetica-Bold-r-Narrow--0-0-0-0-p-0-iso8859-1
hvnbo___.pfb -Adobe-Helvetica-Bold-o-Narrow--0-0-0-0-p-0-iso8859-1
ncr_____.pfb -Adobe-New Century Schoolbook-Medium-r-normal--0-0-0-0-p-0-iso8859-1
nci_____.pfb -Adobe-New Century Schoolbook-Medium-i-normal--0-0-0-0-p-0-iso8859-1
ncb_____.pfb -Adobe-New Century Schoolbook-Bold-r-normal--0-0-0-0-p-0-iso8859-1
ncbi____.pfb -Adobe-New Century Schoolbook-Bold-i-normal--0-0-0-0-p-0-iso8859-1
por_____.pfb -Adobe-Palatino-Medium-r-normal--0-0-0-0-p-0-iso8859-1
poi_____.pfb -Adobe-Palatino-Medium-i-normal--0-0-0-0-p-0-iso8859-1
pob_____.pfb -Adobe-Palatino-Bold-r-normal--0-0-0-0-p-0-iso8859-1
pobi____.pfb -Adobe-Palatino-Bold-i-normal--0-0-0-0-p-0-iso8859-1
sy______.pfb -Adobe-Symbol-Medium-r-normal--0-0-0-0-p-0-iso8859-1
tir_____.pfb -Adobe-Times-Medium-r-normal--0-0-0-0-p-0-iso8859-1
tii_____.pfb -Adobe-Times-Medium-i-normal--0-0-0-0-p-0-iso8859-1
tib_____.pfb -Adobe-Times-Bold-r-normal--0-0-0-0-p-0-iso8859-1
tibi____.pfb -Adobe-Times-Bold-i-normal--0-0-0-0-p-0-iso8859-1
zcmi____.pfb -Adobe-ITC Zapf Chancery-Medium-i-normal--0-0-0-0-p-0-iso8859-1
zd______.pfb -Adobe-ITC Zapf Dingbats-Medium-r-normal--0-0-0-0-p-0-iso8859-1
agw_____.pfb -Adobe-ITC Avant Garde Gothic-Book-r-normal--0-0-0-0-p-0-adobe-fontspecific
agwo____.pfb -Adobe-ITC Avant Garde Gothic-Book-o-normal--0-0-0-0-p-0-adobe-fontspecific
agd_____.pfb -Adobe-ITC Avant Garde Gothic-Demi-r-normal--0-0-0-0-p-0-adobe-fontspecific
agdo____.pfb -Adobe-ITC Avant Garde Gothic-Demi-o-normal--0-0-0-0-p-0-adobe-fontspecific
bkl_____.pfb -Adobe-ITC Bookman-Light-r-normal--0-0-0-0-p-0-adobe-fontspecific
bkli____.pfb -Adobe-ITC Bookman-Light-i-normal--0-0-0-0-p-0-adobe-fontspecific
bkd_____.pfb -Adobe-ITC Bookman-Demi-r-normal--0-0-0-0-p-0-adobe-fontspecific
bkdi____.pfb -Adobe-ITC Bookman-Demi-i-normal--0-0-0-0-p-0-adobe-fontspecific
com_____.pfb -Adobe-Courier-Medium-r-normal--0-0-0-0-m-0-adobe-fontspecific
coo_____.pfb -Adobe-Courier-Medium-o-normal--0-0-0-0-m-0-adobe-fontspecific
cob_____.pfb -Adobe-Courier-Bold-r-normal--0-0-0-0-m-0-adobe-fontspecific
cobo____.pfb -Adobe-Courier-Bold-o-normal--0-0-0-0-m-0-adobe-fontspecific
hv______.pfb -Adobe-Helvetica-Medium-r-normal--0-0-0-0-p-0-adobe-fontspecific
hvo_____.pfb -Adobe-Helvetica-Medium-o-normal--0-0-0-0-p-0-adobe-fontspecific
hvb_____.pfb -Adobe-Helvetica-Bold-r-normal--0-0-0-0-p-0-adobe-fontspecific
hvbo____.pfb -Adobe-Helvetica-Bold-o-normal--0-0-0-0-p-0-adobe-fontspecific
hvn_____.pfb -Adobe-Helvetica-Medium-r-Narrow--0-0-0-0-p-0-adobe-fontspecific
hvno____.pfb -Adobe-Helvetica-Medium-o-Narrow--0-0-0-0-p-0-adobe-fontspecific
hvnb____.pfb -Adobe-Helvetica-Bold-r-Narrow--0-0-0-0-p-0-adobe-fontspecific
hvnbo___.pfb -Adobe-Helvetica-Bold-o-Narrow--0-0-0-0-p-0-adobe-fontspecific
ncr_____.pfb -Adobe-New Century Schoolbook-Medium-r-normal--0-0-0-0-p-0-adobe-fontspecific
nci_____.pfb -Adobe-New Century Schoolbook-Medium-i-normal--0-0-0-0-p-0-adobe-fontspecific
ncb_____.pfb -Adobe-New Century Schoolbook-Bold-r-normal--0-0-0-0-p-0-adobe-fontspecific
ncbi____.pfb -Adobe-New Century Schoolbook-Bold-i-normal--0-0-0-0-p-0-adobe-fontspecific
por_____.pfb -Adobe-Palatino-Medium-r-normal--0-0-0-0-p-0-adobe-fontspecific
poi_____.pfb -Adobe-Palatino-Medium-i-normal--0-0-0-0-p-0-adobe-fontspecific
pob_____.pfb -Adobe-Palatino-Bold-r-normal--0-0-0-0-p-0-adobe-fontspecific
pobi____.pfb -Adobe-Palatino-Bold-i-normal--0-0-0-0-p-0-adobe-fontspecific
sy______.pfb -Adobe-Symbol-Medium-r-normal--0-0-0-0-p-0-adobe-fontspecific
tir_____.pfb -Adobe-Times-Medium-r-normal--0-0-0-0-p-0-adobe-fontspecific
tii_____.pfb -Adobe-Times-Medium-i-normal--0-0-0-0-p-0-adobe-fontspecific
tib_____.pfb -Adobe-Times-Bold-r-normal--0-0-0-0-p-0-adobe-fontspecific
tibi____.pfb -Adobe-Times-Bold-i-normal--0-0-0-0-p-0-adobe-fontspecific
zcmi____.pfb -Adobe-ITC Zapf Chancery-Medium-i-normal--0-0-0-0-p-0-adobe-fontspecific
zd______.pfb -Adobe-ITC Zapf Dingbats-Medium-r-normal--0-0-0-0-p-0-adobe-fontspecific

------------------------------

Subject: gspreview
From: [e-mail]

This is another front-end for ghostscript.  I have gotten and built
it, and actually preferred the user interface, but it had a slight
glitch which was probably me doing a careless install rather than a
program bug.  It didn't seem as feature-full as ghostview, though.
(Not that there are all THAT many features in ghostview, but it does
its job well)

ftp.x.org:/contrib/GSPreview...

------------------------------

Subject: xdvi
From: [e-mail]

A beautifully legible previewing program for dvi with a handy
zoom+pan feature.  Will not interpret ps specials, which are
understood only by dvips (back to the compiler, object file, and now
linker analogy :)  To view a file, do:

xdvi file.dvi

Mine (from ftp.x.org:/contrib/... ) has the selfile feature,
where xdvi puts up a file browser... cute but useless.  This comes
with either TeX or X in the SLS -- I don't remember.  Either way,
you've probably got one.

------------------------------

Subject: xtex
From: [e-mail]

Xtex is similar in purpose to xdvi.  I have tried to build it under
Linux and failed.  It is available as:

ftp.x.org:/contrib/xtex-2.18.5.tar.Z

------------------------------

Subject: gxditview
From: [e-mail]

Produces a preview version of troff source using X fonts (this means
it's legible and typeset-looking.)

groff -TX100 -mandoc man_page.1

will run gxditview to show you a typeset version of the man page.
-TX75 is the same thing, but tiny.  My SLS didn't have a working one
at all.  A good one comes with the source to groff, which you might
want to get anyway for the additional drivers (the SLS, until
recently, was missing several, including postscript).

prep.ai.mit.edu:/pub/gnu/groff-xxxx.tar.z

------------------------------

Subject: non-x-previewing
From: [e-mail]

Gs comes with pc video hardware drivers, but under unix these are not
a good thing.  However, there are gs binaries around which will use
the Linux VGA library (vgalib).  I do not know where to find them or
the patch, but they must be on tsx-11 somewhere.  The ghostscript
device for this is called linux, thus

gs -sDEVICE=linux file.ps

will show you an image of the Postscript.  The environment variable
GSVGAMODE is important for this.  Set it to the nuber of the video
mode you want, taken from the vga.h which comes with vgalib.

At one point the SLS binaries had this driver, then they didn't, and
now they do again (I saw an announcement to this effect Aug 14).  (gs
-help will list the available drivers in your binary.)

Texmgr is a program which will preview dvi under mgr.  I don't beleive
that it currently works under Linux MGR, but if it does, MGR uses
sufficiently less memory and disk that this might be an attractive
option for some.

------------------------------

Subject: ASCII TRANSLATION   <<<<<<------------
From: [e-mail]

These sections describe various programs which can generate plain
ascii from some file formats.

------------------------------

Subject: from TeX
From: [e-mail]

Lametex will generate ascii from tex source.  a beta (?) version is
now available:

sunsite.unc.edu:/pub/Linux/apps/tex/lametex.tar.z

L2a is a program available via FTP which is currently NOT in beta.
However, as lametex is being tailored to the needs of the Linux
community in general and the LDP in particular, I feel it is the
better choice for Linux users.  (But feel free to differ).

------------------------------

Subject: from dvi
From: [e-mail]

Dvi2tty is the name of a program which will process dvi into text.
Aparently, it will also make an effort at reproducing graphics as
well.

------------------------------

Subject: from ps
From: [e-mail]

Ghostscript 2.6.1 comes with a script file which will use gs to
extract just the text from a ps file, called ps2ascii.  See the gs
information above for where it can be found.  Further documentation is
in the ghostscript 2.6.1 distribution files `gs_2asc.ps' and `use.doc'

------------------------------

Subject: from troff
From: [e-mail]

groff -Tascii or -Tlatin1.....

------------------------------

Subject: from ascii/latin1
From: [e-mail]

The GNU recode program handles conversion between various forms of
straight text encoding, ie from Latin-1 to ASCII.  This would be
available on prep.ai.mit.edu.

------------------------------

Subject: CONTRIBUTORS        <<<<<<------------
From: [e-mail]

The following people have offered suggestions, answered surveys,
provided examples, or otherwise contributed to this document:

Steven S. Dick
Ian McCloghrie
Scott Doty
Keith Walker
Werner Almesberger
Hal N. Brooks
Kai Makisara
Kris Gleason
Savio Lam
Matthew Ewin Maguire
James Henrickson
Telly Mavroidis
Lloyd Miller
Richard Brown
Erik Stenvall
Ian Turton
Ian Jackson
Harri Pasanen
Nils Rennebarth
Rusty Carruth
Martin Neimeier
Brian McCauley
Michael K. Johnson
Erik Stenval
Fazal Majid
Tim Thiesen

--------------------------
End of digest
*************