Date:	Wed, 4 Aug 1999 09:45:05 +0300 (IDT)
From:	Eli Zaretskii <eliz AT is DOT elta DOT co DOT il>
X-Sender:	eliz AT is
To:	DJ Delorie <dj AT delorie DOT com>
cc:	djgpp-workers AT delorie DOT com
Subject:	Re: __stub_foo macros in header files
In-Reply-To:	<199908031356.JAA15480@envy.delorie.com>
Message-ID:	<Pine.SUN.3.91.990804094433.6630B-100000@is>
MIME-Version:	1.0
Reply-To:	djgpp-workers AT delorie DOT com
X-Mailing-List:	djgpp-workers AT delorie DOT com
X-Unsubscribes-To:	listserv AT delorie DOT com

On Tue, 3 Aug 1999, DJ Delorie wrote:

> Hmmm... kill() works, there just never is another program suitable for
> tail to monitor.

On DOS, yes; but not on Windows.  "tail --forever" works just fine on
Windows 9X, and even better on NT, where you really *can* rename or
remove an open file, then rename/create it back, and have the expected
Unix behavior.  But `kill' cannot monitor other VMs.

> Wouldn't tail just print "no such process" or something?

The question is: how do you define the condition when such a message
is printed.

Or are you suggesting a change to `kill', for the branch where the
passed pid is not the caller's pid?

> As for the __stub hack, I think it's a hack and I'd rather avoid
> adding hacks to djgpp's headers.

Then how about changing `kill's behavior instead, so this case would
be easily detectable by an application?  Right now, we simply return
non-zero and don't even set errno.

- Raw text -