2 Jun 2002 10:40
Re: kermit and tramp?
Kai Großjohann <Kai.Grossjohann <at> CS.Uni-Dortmund.DE>
2002-06-02 08:40:32 GMT
2002-06-02 08:40:32 GMT
Horsley Tom <Tom.Horsley <at> ccur.com> writes:
>> Does one send a kermit command to the remote end, then send an escape
>> char to Kermit, then tell the Kermit to send (receive) the file? Hm.
>> Maybe the current file transfer mechanism needs a little bit of
>> tweaking to allow this.
>
> I haven't thought about it much, but I sort of envision defining
> "sessions" as kermit scripts. The local emacs would run kermit
> on the local system as a subprocess, feeding it the session script
> argument. The session script would do all the connecting, logging
> in, starting remote kermit, etc with the kermit scripting
> capabilities. It could start the remote kermit in server mode and
> the local kermit script could then use the kermit INPUT command
> to read commands from tramp and act on them.
Hm? This sounds as if you need to implement in Kermit lots of things
that Tramp already does reasonably well. I'm not sure that this is
the right solution.
Tramp executes a lot of different shell commands during normal
activity. Even if you don't use filename completion.
So the only thing I really need is:
(1) How do I get a shell prompt from Kermit?
This goes into a special connection function. Look at
tramp-open-connection-rsh and tramp-open-connection-telnet for
some examples.
Probably it is enough to start Kermit, then send "telnet
(Continue reading)
RSS Feed