>
> [EMAIL PROTECTED] wrote:
>
> > How about using the break condition instead. This is only for the
> > serial port, and most terminal emulators (e.g. kermit, minicom) provide
> > a means to generate a break condition on the serial port.
>
> ---
> As much as I like that idea, I wondering -- if I am "telnet'ed" into
> a serial multiplexor or even a linux box, will breaking into telnet
> command mode and sending a 'telnet brk' get transmitted through the
> current apps/hardware and be translated to a serial-break on the serial
> line. That requires a leap of faith I wouldn't be be comfortable
> making or is usage over telnet-serial "routing hardware" (i.e. Annex boxes),
> "telnet'ing", "rsh'ed", "ssh'ed" to a system and "cu"ing to a serial port not
> considered important?
>
> -linda
Hi Linda,
I don't think it is an issue. No matter how you get there, you'll
eventually have to use 'cu', 'kermit' or an e.g. Annex box to get to
the serial port hardware. It is at the interface to the serial port
itself that the break needs to be generated (I know that the annex
boxes have the capability of generating a break with the reverse telnet
stuff they support). For example, I could have ssh'ed into host
pendragon, telneted to bonnie and cu'ed through bonnies serial port
to kdb on a linux box. I'd just need to ensure that the escape character
for cu (default ~) is different from the escape characters for the intermediate
steps (~ for ssh, ^] for telnet). Likewise if kermit, or minicom,
or even reverse telnet (aka network serial concentrators) are used.
(I believe there is a way to pass the escape character through ssh
if necessary to have the same escape character).
scott
