Ahh we are going to implement STREAMS of some kind in kernel then...
> We would want something similar that sat on the output side
> and changed some output sequences into operations on the console
> display itself.
>
> Of course, you could do much of this as a user program that used
> a pty and wrote directly to /dev/vcsa? devices. Then the kernel
> side could be _really_ light :-).
>
> Mike
Yes! Ever looked at the terminfo record for the sun console
terminal. I think the current linux tty driver does already
far too much in the kernel itself.
The rest should be handled by a pty user land programm.
(Basically I think this should all be done as an extension to the
GNU screen programm.)
I'm even considering the VT switching stuff in kernel
as inherently bad design. Do you remember how much grief this
is all causing since even the beginning of Linux.
Races races races! Security security security!
And Yes I don't care that much about console speed. There is just
a *human* limit to this and most of the currently interactively
used computers are already sufficiently fast in any case.
Thus a clean design over ultimate speed optimization
seems to be well justifyed to me.
The current linux console does scroll already in a speed which
doesn't give me even a chance to get a grasp what's passing by.
I don't consider this as a feature... (Oh FB based console seems to be
much better with respect to this problem :-).
--Marcin
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/