On Fri, May 25, 2012 at 03:54:24PM +0200, Thorsten Kukuk wrote:
On Fri, May 25, Tomas Mraz wrote:
> On Fri, 2012-05-25 at 15:46 +0200, Thorsten Kukuk wrote:
> > On Fri, May 25, Tomas Mraz wrote:
> > > On Thu, 2012-05-24 at 13:59 +0200, Tomas Mraz wrote:
> > > > The attached patch changes the semantics of the default (non
enforcing
> > > > for root) behavior of pam_pwhistory more reasonable and aligned with
the
> > > > default behavior of pam_cracklib. It will not enforce error but will
> > > > print a message in case the root changes password to a duplicate
value
> > > > for himself or for another users. The behavior when the option
> > > > enforce_for_root is used stays unchanged.
> > > >
> > > > OK to commit?
> > >
> > > Any comments?
> >
> > Since the info for root is only informativ and no error, if
> > it is not enforced for root, we shouldn't use pam_error.
> > This will only confuse people: getting an error message but
> > without the knowledge that the passwort has still be changed.
>
> I'll replace pam_error in the non-enforcing path with pam_info.
> Otherwise is it OK?
Yes, else it is fine with me.
Fine with me, too.
--
ldv