On Wed, 19.12.07 11:32, Nicolas Mailhot (nicolas.mailhot(a)laposte.net) wrote:
> The right behaviour is to make sure inactive sessions cannot
interfere
> with your IO devices while they are inactive. And IO devices are
> keyboards, mice, screens *and* sound cards.
Sound cards are *not* like keyboards, mice and screens.
A keyboard is only providing a service to the person typing on it
A mice is only providing a service to the person which has it in its hand
Any semi-decent speaker can provide music services to everyone in a
room, (even a different room of the one where the computer is) not
just the person in front of the computer. Time shifting means a user
can access sound devices even after his desktop session went inactive
or closed.
Printing is outputting info too. Will we limit printer access to the
active session too to get the "right" behaviour?
With a printer you cannot really spy on people. With access to the
sound card you can. Very easily.
Lennart
--
Lennart Poettering Red Hat, Inc.
lennart [at] poettering [dot] net ICQ# 11060553
http://0pointer.net/lennart/ GnuPG 0x1A015CC4