Message-ID: <3957C1E5.DFC7F995@softhome.net>
Date: Mon, 26 Jun 2000 22:49:41 +0200
From: Laurynas Biveinis <lauras AT softhome DOT net>
X-Mailer: Mozilla 4.73 [en] (Win98; U)
X-Accept-Language: lt,en
MIME-Version: 1.0
To: djgpp-workers AT delorie DOT com, Eli Zaretskii <eliz AT is DOT elta DOT co DOT il>
Subject: Re: Patch: chown() preparation for symlinks
References: <Pine DOT SUN DOT 3 DOT 91 DOT 1000626151341 DOT 7682E-100000 AT is>
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Reply-To: djgpp-workers AT delorie DOT com

Eli Zaretskii wrote:
> 
> On Mon, 26 Jun 2000, Laurynas Biveinis wrote:
> 
> > As I can see, all the complexity there is for the case when _chmod()
> > fails, for root directories and device files. So the performance loss
> > should be pretty small for real applications.
> 
> The complexity will come into play each time _chmod fails, i.e. for each
> missing file.

Oh, of course, missing files.

> > BTW, should chown() succeed for those root directories and device
> > files?

> Devices seem especially questionable.  As for root directories, I doubt
> that a Unix program would ever want to call chown on it.

Devices do have owners and they can be freely changed in unix; so I see
it as a bug in DJGPP. I'm not sure about / directory there. 

What's the best way to fix device handling? Make __file_exists() recognize
devices too?

Laurynas