Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!watmath!clyde!burl!ulysses!allegra!mit-eddie!genrad!panda!talcott!harvard!cmcl2!rna!kc
From: kc@rna.UUCP (Kaare Christian)
Newsgroups: net.bugs
Subject: Pic advice (almost a bug)
Message-ID: <479@rna.UUCP>
Date: Mon, 17-Mar-86 23:10:51 EST
Article-I.D.: rna.479
Posted: Mon Mar 17 23:10:51 1986
Date-Received: Wed, 19-Mar-86 05:31:26 EST
Organization: Rockefeller Neurobiology
Lines: 29

Here is my latest pic discovery.  Although this is not a bug
report per se, the following advice may prove very helpful if you
want your pic-tures to come out right.

My version of the pic distribution is setup for the Linotron 202.
One subtle consequence of this is the following. (Note none of this
is #ifdef'd for the lino202.) In an apparent attempt to make pic
usable with the old troff, pic draws short lines using the new
line drawing commands, and draws longer lines using repeated base line
rule characters.  In the file pltroff.c there is a printf
statement that elicits a troff vertical motion before outputing a
string of base line rules. Presumably this vertical motion makes
the base line rule get drawn in just the right place. You might want
to adjust this vertical motion to match the actual placement of the
base line rule for your font/printer. The suplied value is .25m, but
for my QMS laser printer the value zero works better.

Alternatively you can suppress output of horizontal lines using
the baseline rule character, and go with horizontal lines drawn
with the line drawing primitives as built into the newer (di)
troff. You can do this by setting the variable 'useDline' to a one
near the top of the pltroff.c file.

Either way, if you get boxes and lines that don't quite match up,
look to the pltroff.c file for the answer. Tho this isn't a bug,
these printer dependencies could be handled more gracefully.

Kaare Christian
cmcl2!rna!kc