Subject: Re: why must exiting processes wait forever for events that never
To: Paul A Vixie <>
From: David Brownlee <>
List: tech-kern
Date: 11/16/2000 20:41:28
	Continuing the late trend...

On Sat, 28 Oct 2000, Paul A Vixie wrote:

> Sorry I'm late following up on this...
> > > Actually, what makes Mike's solution sillier is that there's already
> > > another way to do it.  `stty -f /dev/tty... -crtscts -cdtrcts -mdmbuf'
> > > should unstick the tty and cause any output to drain (unless the line
> > > has been disconnected, in which case output is flushed anyway).
> That's true if (1) one is aware of every possible kind of flow control NetBSD
> understands, and (2) none has been added since the last time one read stty(1).
> Neither of those is true in my case.  I believe Mike implemented "stty
> flushout" in terms of the options shown in the above snippet.  Because I
> want on some days to just be a user, I want commands to describe what I
> am trying to do rather than what bits to turn on or off to get it done,
> and I do not think NetBSD ought to require relatively deep technical
> knowledge in order to get the least-astonishing behaviour.
> And in that last vein, a timeout in close() makes way more sense than
> "stty flushout", just as "stty flushout" makes way more sense than
> "stty -crtscts -cdtrcts -mdmbuf".

	Does "stty flushout" reset the flags after the output has drained
	(presumably some arbitrary short time), or does it rely on that
	happening on close?

			       -- A pmap for every occasion --