X-Authentication-Warning: delorie.com: mailnull set sender to djgpp-workers-bounces using -f
Date: Thu, 27 Dec 2001 16:37:49 +0200 (IST)
From: Eli Zaretskii <eliz AT is DOT elta DOT co DOT il>
X-Sender: eliz AT is
To: djgpp-workers AT delorie DOT com
Subject: Re: lfn from scratch...
In-Reply-To: <200112271353.OAA16101@father.ludd.luth.se>
Message-ID: <Pine.SUN.3.91.1011227163127.21698A-100000@is>
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII
Reply-To: djgpp-workers AT delorie DOT com
Errors-To: nobody AT delorie DOT com
X-Mailing-List: djgpp-workers AT delorie DOT com
X-Unsubscribes-To: listserv AT delorie DOT com
Precedence: bulk


On Thu, 27 Dec 2001, Martin Str|mberg wrote:

> > E.g. try "ls \\nonexistent\machine" some day.
> 
> Yes. And your point is?

The point is that "ls nonexistent-file" returns immediately, as does
"ls x:" for a non-existent drive x:, while the above waits for a long 
time before it returns to the caller.

Thus, if some program tries a list of possible directories until it 
succeeds to find a valid one, it would work much slower, and could even 
appear to hang, if one of the directories references a UNC, instead of 
failing immediately.

One program which tries such a list of directories is GCC.

> (How would ever make _that_ work?)

I didn't mean it to work, but I do want the failure to be fast.