Date:	Wed, 28 Jun 2000 16:11:17 +0300 (IDT)
From:	Eli Zaretskii <eliz AT is DOT elta DOT co DOT il>
X-Sender:	eliz AT is
To:	Laurynas Biveinis <lauras AT softhome DOT net>
cc:	djgpp-workers AT delorie DOT com
Subject:	Re: Patch: chown() preparation for symlinks
In-Reply-To:	<3959BF50.6FB7BB3E@softhome.net>
Message-ID:	<Pine.SUN.3.91.1000628160907.526D-100000@is>
MIME-Version:	1.0
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

On Wed, 28 Jun 2000, Laurynas Biveinis wrote:

> Eli Zaretskii wrote:
> > If we want chown to support devices and root directories, you might
> > as well use your original code with access instead of __file_exists.
> > It doesn't make sense to add complexity to __file_exists just to make
> > a no-op function such as chown be marginally simpler.
> 
> So __file_exists() is intentionally meant not to detect devices?

Yes, __file_exists was written to be lean and mean, only for files,
it is meant to avoid all the complications of Posix compliance.

> This sounds a little bit dangerous to me. At least it should be
> documented.

Yes, documenting this would be a good idea.

> But in any case, chown() should use access(). OK to commit?

Yes, I think so.

- Raw text -