> Indeed. X does have the opportunity to mess with the keys before
> they're passed to wscons. For example, on the workstation I'm in
> front of now, CONTROL_R/ALT_R/F1 switches to the console, but wscons
> only recognizes *left*-CTL/*left*-ALT/F5 to switch back to X. I don't
> suppose your X setup remaps any of the keys?
i don't think my wm (fluxbox) does it. there's alt-f[1-4] function to switch the 
workspace, but nothing connected with ctrl... when a terminal (aterm) with 
midnightcommander is focused, pressing ctrl-alt-f1 prints [11^; in plain aterm 
it prints nothing, however it then doesn't react on keyboard input (except 
<enter> which brings new prompt)...

i just made some tests - here are the results

1) commented out screen 0, consdev pc (default)	switching ok
2) commented out screen 0, consdev com0		X switching no, commandline ok
3) uncommented screen 0, consdev pc		switching ok
4) uncommented screen 0, consdev com0		switching ok

so it seems that it requires screen 0 uncommented. is it a consdev bug (btw, i 
use consdev auto, not consdev com0, but the output goes to com0)?

i've got another question concerning using setup 3). i uncommented ttyE0 in 
/etc/ttys, but i cannot login from ttyE0/console (it's reported as console). it 
asks for login and then password, but does nothing. after i press the enter key 
for the second key it returns "login incorrect". after another login attempt, 
the prompt for password is not displayed at all. login from ttyE[1-3] works 
fine. i thought the problem could be caused by not matching values of the type 
field in wscons.conf for console and ttyE0 (vt100 vs. vt220) so i changed ttyE0 
to vt100, but it had no effect. in fact, there's such mismatch for all ttyE* 
between /etc/ttys file contents (all ttyE* are vt220) and wscons.conf (all ttyE* 
are vt100).
btw, i haven't seen any "... already configured" message during the startup.

when serial console is attached (case 4) everything works fine.

-- 
Lubos
_@_"