Pavel:

>As you discovered, the kernel detects what device to use as a 
>console and presents it as /dev/console. That means that ttyB1 
>should not be used, because there would be two gettys on the same 
>physical port which would not work. That's why there is a comment:
>
># In normal operation, getty on 'console' pseudo-device is disabled,
># and is enabled for the correct 'real' device (e.g. one of the serial
># ports).
>
>Don't do this except for testing, because (at least on my DEC3000 -
>600) I observed that the ttyB1 device stopped working in -current. I
>am actually curious if it is true for your machine too.

For reference, when I tried on my 3000/400 with this line in 
/etc/ttys commented out

   console "/usr/libexec/getty std.9600"   vt100 on secure

and ttyB0's line changed to:

   ttyB0  "/usr/libexec/getty std.9600"   unknown on secure

this is where the console gets to before stopping:

Setting securelevel: kern.securelevel: 0 -> 1
Starting virecover.
Starting local daemons:.
Updating motd.
Starting sendmail.
Starting inetd.
Starting cron.
Sat Sep 11 16:28:58 EST 2004


with ttyB1's line changed to

   ttyB1  "/usr/libexec/getty std.9600"   unknown on secure

the console stops responding here:

Setting securelevel: kern.securelevel: 0 -> 1
Starting virecover.
Starting local daemons:.
Updating motd.
Starting sendmail.
Starting inetd.
Starting cron.
Sat Sep 11 16:32:36 EST 2004
Sep 11 16:32:39 jkalpha2 getty[342]: /dev/ttyB1: Device not configured



Ray