Mailing-List: contact cygwin-help AT sourceware DOT cygnus DOT com; run by ezmlm List-Subscribe: List-Archive: List-Post: List-Help: , Sender: cygwin-owner AT sources DOT redhat DOT com Delivered-To: mailing list cygwin AT sources DOT redhat DOT com Date: Wed, 21 Feb 2001 21:26:56 -0500 From: Christopher Faylor To: cygwin AT cygwin DOT com Subject: Re: [ANNOUNCEMENT]: Important change to symbolic link functionality Message-ID: <20010221212656.F7576@redhat.com> Reply-To: cygwin AT cygwin DOT com Mail-Followup-To: cygwin AT cygwin DOT com References: <20010221232921 DOT X908 AT cygbert DOT vinschen DOT de> <000a01c09c73$7a43ce10$d938a8c0 AT Hadfield> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline User-Agent: Mutt/1.3.11i In-Reply-To: <000a01c09c73$7a43ce10$d938a8c0@Hadfield>; from m.hadfield@niwa.cri.nz on Thu, Feb 22, 2001 at 03:02:14PM +1300 On Thu, Feb 22, 2001 at 03:02:14PM +1300, Mark Hadfield wrote: >OK, so I can avoid this by using Cygwin, right? Well no. Doing a "ls" on a >directory with migrated files is OK, doing an "ls -l" takes several minutes, >I think because ls tries to read data from every file to see if is a >symlink. Command completion also seems to involve looking up symlinks, so is >unusable. No. Cygwin only does symlink tests on files which have the system attribute set. This was hashed, rehashed and bitterly noted in a recent thread here. It does open the file to see if it has a '#!', marking it executable if so. Possibly setting CYGWIN=ntsec will avoid this overhead. Mounting the directory with the '-x' flag will certainly avoid it at the expense of having cygwin think that every file in the directory is executable. >If Cygwin were to do its symlink checks only on files with the .lnk >extension, then these problems would go away. > >It has been on the back of my mind for a while that symlinks are supposed to >have their system bit set. As far as I can tell, none of the files on the >DMF area, as served by Samba, does have its system bit set. So "ls -l" >shouldn't be looking inside any of the files to see if they're symlinks. >Perhaps there's some other reason to look inside the file? Oh well, I guess >that's one of life's little mysteries I'll never solve. But the new symlink >functionality sounds like a good idea anyway. You could solve this by either reading the code or paying attention to the mailing list... cgf -- Want to unsubscribe from this list? Check out: http://cygwin.com/ml/#unsubscribe-simple