I would like to request comments and suggestions on the following plan:
I am attempting to use the net as a substitute for long-distance dedicated
lines. I have terminals in one city that need to talk to a mini-computer in
another city. The mini-computer has NO internet access at all. I am planning
on two freebsd machines, one with the terminals hooked up to it, and one
with multiple serial ports connected to the mini-computers RS-232 lines. The
shell script for each user would directly start up a TELNET process, and log
them off when TELNET ends. They will not be doing ANYTHING else on the system.
Once a connection is made, I need ALL characters to pass through from the
terminal to the mini, and back, WITHOUT INTERPRETATION.
How do I get the local process on the terminal side to ignore everything? I
looked at setting the tty to non-canonical, and BRKINT, so that a BREAK
could shut down the telnet process and log them off. Is this correct?
On the machine by the mini, it appears that:
1. adding the line #define _PATH_LOGIN 'my-prog'
2. recompiling TELNETD with another output name
3. putting this new program in /etc/services on an unused port
4. specifying the new port to TELNET on the terminal machine
would let me get from a terminal to the pty created by TELNETD. Is this correct?
'my_prog' would be a program that looks for the first available serial port,
and then connects serial-in to pty-out and serial-out to pty-in. Is there an
existing program that might do this? or that I could use code from? Again, I
need ALL characters to go thru as is.
Thanks in advance for your help. I am just getting started in freebsd, it is
just starting to make sense, and that scares me a little.
Randy Berndt
[Signature, like every damned freeway in Houston, is "Under Construction"]