Sean Middleditch wrote:
Yes, we do use kickstarts... But that isn't really feasible on the laptops we are installing, because the hardware is different on nearly each one and everyone wants to be root on their laptop. Having root on their laptop means, of course, that the laptop isn't trusted on our network...Kim B. Nielsen wrote:
Actually, for my work, I use the graphical root login for one thing:
Setting up a freshly installed machine... I know that this can be done from an user account, but I
Kickstart doesn't help? It's a fairly useful technology that only Red Hat has a good implementation of. (That I'm aware of.)
We are, however, using kickstarts to a large extent on our regular machines.
Another issue is, that a system, at install time, can be configured without any users at all. This is for instance the case if you configure network authentication. If this setup for some reason doesn't work, and root cannot log in (graphically), the user is left with a system he cannot login to. Surely he can use a graphical console, but often a graphical login will be more usefull, since it can be easier to fix thing. Anyway, you are able to have a webbrowser and a console on the same screen, which I have found to be very usefull :)
But an admin can get into X even if GDM disallows it. And only an admin is setting up machines with network authentication. It's the home users that don't have a fricken' clue that think, "oh, I'll log in as this root thingy, it's soooo much easier!" that are the problem.
True... But I think that the doomsday screen and warning message will work better, than disallowing the graphical login. When you are setting something up on a machine, it's sometimes easier to use the graphical root to configure the thingy and make it work, and then, afterwards, figure out how to make normal users use the thingy.
However, I'm voting in favor off a warning message at log in, and possibly some doomsday looking desktop.
Regards /kbn
-- "He who controls the proxy server, controls the world..."
