urban@Ucb-C70@sri-unix (11/23/82)
From: randvax!trw-unix!trwspp!sentry1!urban at Ucb-C70 (Mike Urban)
To: trwspp!trw-unix!randvax!ucbvax!unix-wizards at Ucb-C70
Via: Ucb-C70; 20 Nov 82 16:47-PST
Remailed-date: 22 Nov 1982 1833-PST
Remailed-from: the tty of Geoffrey S. Goodfellow <Geoff5 at SRI-CSL>
Remailed-to: Unix-Wizards@SRI-CSL: ;
Via: Sri-Csl; 23 Nov 82 0:35-PST
Date: Friday, 19 Nov 1982 09:16-PST
(This message doesn't seem to have made it the last time I
sent it).
At TRW, we are experiencing several problems with "tbl".
Most of the problems involve tables that span more than
one page break; tables with a large number of text
diversions also seem to cause strange problems. Some
examples:
o Boxed tables cause errors both in nroff and troff when
more than one page in length.
o The message "cannot do .ev" is sometimes produced by
Nroff (-me) on large tables. This happens with a -T
option for some devices but not others.
o Tbl isn't well bulletproofed against naive-user
errors, like a column with no entries. Some of these
cause core dumps.
o Some of our users have managed to write tbl tables that
cause nroff to generate Spinwriter codes resulting in table
entries whose baselines descend fractionally between
the beginning and end of the entry (does this make
sense? I mean that the table entry, when printed,
"goes downhill" on the paper). This messes up the
effective page length.
The system in question is a Vax 11/780 under 4.1bsd.
Is there someone out there who has a version of "Tbl" that
actually works reliably for large tables? How about an annotated
version of the source code? How about a version of Tbl that produces
stuff for TEX? Anything. Help......
Mike