From: sandmann AT clio DOT rice DOT edu (Charles Sandmann)
Message-Id: <10109261602.AA14411@clio.rice.edu>
Subject: Re: fixpath patch (fixes rm -rf disaster, retains deep directory
To: wojciech DOT galazka AT polkomtel DOT com DOT pl (Wojciech Galazka)
Date: Wed, 26 Sep 2001 11:02:18 -0500 (CDT)
Cc: djgpp-workers AT delorie DOT com ('djgpp-workers AT delorie DOT com'),
   eliz AT is DOT elta DOT co DOT il (Eli Zaretskii)
In-Reply-To: <250B3114DA16D511B82C00E0094005F8023FC1A5@MSGWAW11> from "Wojciech Galazka" at Sep 26, 2001 04:23:43 PM
X-Mailer: ELM [version 2.5 PL2]
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
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

> > We cannot simply replace getcwd with _truename here: _truename returns
> > a UNC for directories on networked drives, and some DOS calls on some
> > supported systems, notably on plain DOS, don't support UNC's.
> > 
> > _fixpath uses getcwd precisely for this reason.
> 
> I think we should watch int 21h, ah = 3bh (set current directory) and if the
> path is larger than 64 characters then return error 

That was my first patch, it fixed some problems but did not fix the root
problem, which was fixpath.

We need the ability to use deep paths for complex builds on Windows 2K.
All other calls seem to behave properly if we modify fixpath to avoid using
a bad getcwd.

Maybe we should only use _truename if we return root as the path.
I also did not try using the findfirst call in this patch, but also seemed
to perform the same action as truename on "." if it would be a better
choice?