Cameron Simpson wrote:
| >Is this the way it should be?
|
| I believe so... if the tail did not subsequently try to show any more
| output. Only when it tries to issue more output to the now gone-away
| terminal session would it get a signal that would end up killing it.
Um, I'm not so sure.
A foreground tail -f should get a SIGHUP and quit. A background one may
behave as you describe, being in a different process group.
Mike: were your "tail -f"s foreground or background?
Hum I just tested the thing on a Konsole of its own, as Cameron says the
tail (on a static file) died straight away.
I thought I had seen the behaviour I described if I killed a remote ssh
session but no that dies too. Maybe it was just things that weren't
going to respond to a signal.
Maybe it can help to strace the tail with 2>/tmp/dump on the end and
have a look in there after killing the konsole?
-Andy
