BJ> I'm just starting to get things together under Linux, including 
 BJ> BinkleyTerm XE.  I've been running BinkleyTerm 2.60 
 BJ> under OS/2.  In that version, there are three 
 BJ> methods for handing off to a BBS, in my case 
 BJ> Maximus/2, version 3.01.  One is via a shell script 
 BJ> error level exit.  This is NOT recommended under 
 BJ> OS/2 (or Win32), and probably not recommended in 
 BJ> Linux.  If I remember correctly, the one I use is 
 BJ> the SpawnBBS method, where Binkley starts up the 
 BJ> Max process directly, passing commandline info, 
 BJ> including (under OS/2) the needed handle to the 
 BJ> serial port.  [Under DOS, this passes the fossil 
 BJ> port number.]  I don't remember the third type at 
 BJ> the moment.  
 
Any idea why errorlevel-exit isn
t recommended? It doesn't seem like such a horrible way to do things,
although I would personally prefer spawning the BBS directly from the
mailer (assuming the mailer doesn't try and diddle the com port into the
controlling tty or something equally annoying). 
 
How does Bink XE handle the errorlevel exit from Maximus? I have a
"jury rigged" solution in testing right now to "handle"
errorlevel exits from the forked maximus child (after accepting a TCP/IP
call) which actually re-calls the runbbs.sh shell script and passes it its
errorlevel on the command line before dying. This allows me to use a single
script to control the ip inbound calls and the serial inbound calls -- and
have the serial calls handled as traditionally as possible.
 
Wes

--- Maximus/2 3.01