>        'gdb /usr/bin/login login.core' and type 'where' to find
>        out what was happening when it dumped core.
>        (I'm pretty sure you need a core from the -g compiled binaries :)

gdb /usr/bin/login login.core:
(no debugging symbols found)...
Core was generated by `login'.
Program terminated with signal 3, Quit.
#0  0x10005369 in ?? ()
but that's w/0 the recompile
(just compiled the sommerfeld-com-961026.tar.gz in, a real 386 is slow)
& didn't get any more .core files

>> and in my config i've got:
>> pseudo-device   ppp     8               # PPP
>> pseudo-device   tun     2               # network tunneling over tty
>>
>> must i highten tun to 8 as well? or is it not related?
>>
>        I'm pretty sure its unrelated.

OK

>        I'm pretty sure port-mac68k is the odd one out in having the
>        ability to have 'termios defaults' for a given port.

i was scared about that too..( main reason to swap ;)' )
but now i don't know what's right and what's wrong
not having a ok situation at hand on the i386

>        If you're running getty on a port then editing /etc/ttys
>        (& kill -1 1) should be enough to have the port running at

strange is that a kill -1 1 doesn't do the job.

>        the right speed (despite what stty -f may say). If you're

i can't rely on stty -f !?
any other i can use to check??
(from the unix side..not that i have to attach a terminal)

>        not running getty then... apologies for missing the start

jep getty's running
  377 ??  IN     0:00.17 /usr/libexec/getty std.57600 tty00
  379 ??  IN     0:00.17 /usr/libexec/getty std.115200 tty03
  382 ??  IN     0:00.17 /usr/libexec/getty std.9600 tty01
  386 ??  SN     0:00.17 /usr/libexec/getty std.57600 tty02

???,
~n