[net.bugs.4bsd] vi doesn't work right on VT100s

guy (02/14/83)

OK, but:

1) There are already programs which misuse "termcap"; the games "rain" and
"worms" are two offenders.  (They send "termcap" strings to the terminal
directly through "fputs" instead of through "tputs"; this means that on a
VT100 you get "5"s all over the screen due to "fputs" trying to put the
delay strings out as text.)  As such, I assumed that the problem was with
"vi", not with "termcap", and that the change to "termcap" had been made but
nobody got around to updating "vi" or the documentation.  (Out-of-date
documentation is, unfortunately, still a way of life in the UNIX community.)

2) The use of "al" and "dl" for the VT100 does make inserting multiple lines
of text, and deleting lines, a bit nicer.

3) Given the number of different ways that terminals handle settable scrolling
regions, can "termcap" and "vi" handle them all?  (At our site, we have a
terminal (Zentec 8001 with custom firmware) where a scrolling region has no
effect on cursor positioning, a terminal (Concept 108) where cursor positioning
is ALWAYS relative to the scrolling region, and a terminal (VT100) where it
can work BOTH ways depending on whether the terminal is in "absolute" or
"relative" cursor positioning mode.)  Furthermore, does "termcap" properly
handle terminals which do NOT have the save/restore cursor capabilities?

4) Given 2), it might be nice to submit the appropriate changes to "termcap"
and "vi" over the network.

					Guy Harris
					RLG Corporation
				(decvax!duke!mcnc!rlgvax!guy)