Subject: tty_login, tty_logout (was: pcvt and TIOCCONS)
To: None <email@example.com>
From: Gordon W. Ross <firstname.lastname@example.org>
Date: 04/19/1996 14:40:33
> Date: Fri, 19 Apr 1996 17:22:25 +0200 (MET DST)
> From: Olaf Seibert <email@example.com>
> Chris G Demetriou <Chris_G_Demetriou@UX2.SP.CS.CMU.EDU> wrote:
> > * check_logger
> > where that action may, say, want to check for logins by
> > individuals suspected of doing Bad Things, etc., and set up some type
> > of logging mechnism, or something...
> And of course, it should be possible to reject a login altogether,
> say if you don't want somebody logging in on a certain port, from
> a certain host, at a certain time, whatever (all at the discretion
> of the called external command of course). Perhaps the program should
> get fds for the terminal (and the network connection, if any) so
> that it could test or manipulate them.
What is the justification for trying to use this "fbtab replacment"
as an extension for new authentication controls? If you want to
have more control over how authentication is done, I think that
should be dealt with separately from the "fbtab replacment" on
the grounds of orthogonality. (Unless somebody can show a good
reason why they should both be done by one mechanism - doubtful.)
Fancy authentication is typically done by replacing login with a
new version supporting the Fancy extensions. That seems quite
appropriate given that login is responsible for that job.
Further, if we check the return status of commands run by this
"fbtab replacment" I submit that commands that return non-zero
status should NOT prevent a login on that line, but merely cause
a warning message to be logged. This will prevent misuse of this
mechanism for authentication, and will avoid situations where you
can not login after having messed up your tty chown script.