Subject: | |
From: | |
Reply To: | |
Date: | Thu, 29 Mar 2001 09:29:52 -0800 |
Content-Type: | text/plain |
Parts/Attachments: |
|
|
Gary writes:
> Question: is it possible for the Break instruction be recorded on another
> machine and sent ?
> is the Break instruction de-codable to its ESC + xxxx value ?
> Thankfully there is an Escape key !
But unfortunately <BREAK> is not an escape sequence :-)
The BREAK signal is an out-of-band indication and as such has no character
equivalent, and it is impossible to generate via the data stream itself.
Your client program must provide some function (a <BREAK> key, menu option
or something else) that will initiate sending the BREAK signal.
On a serial connection, a BREAK is indicated by (IIRC) holding the Transmit
Data line in the "mark" state for at least 200ms, a condition which cannot
occur during normal transmissions.
On a VT or telnet connection there is obviously no Transmit Data "wire"
per-se, but the VT and telnet protocols have places to indicate that a BREAK
is being requested. Again these are out-of-band indications so you can't
just "type" them (unlike CTRL-Y which is an example of an in-band indication
as it is just another character as far as the data stream is concerned).
Fortunately the Telnet protocol, through it was not designed originally for
MPE, does include the out-of-band BREAK concept. For most Unix-like telnet
clients, you can hit CTRL-] when you are connected to a 3000 and at the
resulting 'telnet>' prompt you can type 'send break', and the 3000's telnet
server will properly interpret the signal as requesting the MPE <BREAK>
function.
I note however that the Microsoft telnet client (at least Win2K) does not
seem to support the 'send' command though.
If CTRL-] doesn't work for you, you might want to look at the documentation
for your client and see if it provides some other keystroke combination to
access this functionality.
G.
|
|
|