I telnet in to either d3/linux or d3/nt each day;
at the end of the day I do EXIT from d3/linux,
as that drops the telnet session and frees up
the dynamically-assigned pick port number.
But I do OFF from d3/nt at the end of the
day since EXIT doesn't seem to drop the
telnet connection. Is avoiding EXIT from
d3/nt ill-advised?

From our experience using D3/NT:

Using "exit" on a dynamically assigned telnet session frees up the port.
Using "off" leaves the port open to log on again.

Using "exit" on a serial port will remove (dev-remov) that port, so always
use an "off".
To use the serial port again after an exit you would have to do a dev-make.

And I would think using an "exit" on a nailed telnet port would remove that
port, use an "off".

Frank Winans wrote:

Craig Taylor Equipment ha scritto:
need to do a dev-make

Frank:

I use "OFF" in both platforms. When I look at the NT connections, via
NT_PINFO, the ports are freed up. On Linux a "ps aux" shows the PID also
removed (mostly). If the PID isn't removed a "PID *" will display the Linux
PIDs attached to D3 ports. If one exists you want to get rid of you can
issue the D3 command "KILL P{n}" where "{n}" is the D3 port# to kill the
underlying Linux PID for.

I have the line "DCD (NS" in each user's macro (in both NT and Linux) and
the line "TRAP DCD EXIT" in the "user-coldstart" of Linux only (it doesn't
seem to matter with NT on my laptop).

Hope this helps.

Bill

"Frank Winans" <fwinans (AT) sbcglobal (DOT) net> wrote

Bill,

Do the same as you and also added a timeout to most users macros.
Also have copied exit and off to :exit and ff in the md, then written a
exit and off bp program
to check ports to determine if I really want to exit (:exit) or logoff
(ff).
This worked good both on D3/NT and D3/SCO.
Never do an exit on a serial port on D3/NT, it will do a dev-remov of that
serial port.

Bill H wrote: