| Date Of Receipt: 15-MAR-1996 12:08:51.43
From: SMURF::FLUME::jmf "Joshua M. Friedman OSF/UNIX SDE 15-Mar-1996 1205"
To: kucherov@DEC:.zko.flume
CC: odehelp@DEC:.zko.flume, tresvik@DEC:.zko.flume
Subj: re: Bug report for srequest
I'd say the majority of people use vi for their srequest edits. This is
quite an odd situation that you'd ^C in the middle and expect to continue
the srequest. I'd say normally the only reason to ^C is because you don't
need any edits, in which ":q" or ":q!" would be the most logical choice.
We may look into a fix for this but I'd say it's very low priority
compared with other pending work.
-josh
------- Forwarded Message
To: odehelp
Cc: kucherov
Subject: Bug report for srequest
Date: Wed, 13 Mar 96 13:00:36 -0500
From: kucherov
X-Mts: smtp
If you use vi as your editor (not recommended due to this bug),
srequest will abort when you do ^C, and you'll have to start
all over. srequest should not be receiving signals from its child, vi.
srequest should, as a first cut, simply ignore ^C interrupts
while vi is executing.
This would be a huge improvement over its current behavior.
sergei
P.S. After a ^C in vi, srequest kicks you out of vi (you lose)
and says: Send this form [y|n] ?:
P.P.S. Other common software has a similar bug, such as the
elm mailer. Years ago I fixed a similar bug in the curses-based
program called "submit" (Ultrix tool for doing submits -- I fixed
the MLS version of it).
Lot's of folks get this simple signal handling wrong.
Users such as myself should just not bother using vi with these
front-end tools that misbehave vis-a-vis catching signals.
------- End of Forwarded Message
|