[3153] in linux-net channel archive
Re: gated 3.6A1 for 1.3.99 (diffs)
daemon@ATHENA.MIT.EDU (Marek Michalkiewicz)
Sun Jun 9 16:39:15 1996
From: Marek Michalkiewicz <marekm@i17linuxb.ists.pwr.wroc.pl>
To: gafton@sorosis.ro (Cristian Gafton)
Date: Sun, 9 Jun 1996 15:18:53 +0200 (MET DST)
Cc: scldad@sdc.com.au, linux-net@vger.rutgers.edu
In-Reply-To: <Pine.LNX.3.91.960609150446.6124B-100000@main.sorosis.ro> from "Cristian Gafton" at Jun 9, 96 03:10:37 pm
Cristian Gafton:
> Still, I seriously doubt that this will work on shadow-aware systems
> without actually having /etc/shadow.
Why not? Many other programs do it this way and work just fine -
if /etc/shadow doesn't exist, getspnam() will return NULL and the
password from /etc/passwd will be used. If you see any specific
problems with this, let me know.
> Why not check the presence of /etc/shadow and if it is present and size >
> 0 then proceed with shadow things ?
Because it's more code which is not necessary in most cases. I do it
only in programs which need to modify password files (passwd already
works with both shadow and non-shadow passwords in my current, not yet
unreleased source).
Besides, the implementation of getspnam() might change later, so there
shouldn't be too many programs with hardcoded check for /etc/shadow.
(Not that it is expected to change anytime soon, but...)
Regards,
Marek
