> [ On Fri, January 22, 1999 at 08:44:30 (EST), Jeffrey Altman wrote: ]
> > Subject: Re: How do I sz/rz under an ssh connection?
> >
> > > > I can and has been for many years.
> > > >
> > > > SET HOST * <port>
> > >
> > > Hmm... I didn't know about this, and it's not given as an example in
> > > the on-line help. I was looking for a command-line option on C-Kermit
> > > to tell it to listen on a specified port (presumably in "server" mode),
> > > but I didn't see one....
> >
> > All commands can be issued on the command line with the -C "<command list>"
> > option.
>
> Sure, but that still didn't help me in my quest to find some way to get
> kermit to listen for a TCP/IP connection on a given local TCP port.
Huh?
kermit -C "set host * <port>"
what is so hard about that?
> > > > SET HOST <host> <port>
> > >
> > > It should be possible then to tunnel a kermit connection through SSH,
> > > though I've never set SSH tunnels up, so I don't have a cook-book
> > > example of how this might be done...
> >
> > Right. But if 'ssh' did not require the controlling process to be a
> > PTY I could simply say:
> >
> > SET NETWORK TYPE COMMAND
> > SET HOST ssh host
> >
> > But because 'ssh' requires that the controlling process be a PTY for
> > it to make a terminal connection this does not work.
>
> That's not what I was thinking.
>
> There should be no need for kermit to provide a controlling process for
> SSH if you use an SSH tunnel.
Right, but if I use my format than I do not need to use a tunnel which
is very unintuitive for most end users.
> Assume I've set up an SSH connection to a remote host, and I've told SSH
> to forward any connections to local port 22222 through to the remote
> port 11111.
>
> I should then be able to set up a kermit server process listening on
> port 11111 on the remote system:
>
> kermit
> > set host * 11111
> server
kermit -C "set host /server * 11111"
> (of course I won't see the prompt and the entry to server mode until a
> connection is completed, but that's OK as it seems to at least accept
> type-ahead -- all that's missing is some scripting to re-start the
> listen after a connection closes so that you can use as many kermit
> sessions as you want until you interrupt the kermit server script)
kermit -C "while true { set host /server * 11111 }"
> Then on the local system I need only type "kermit -j localhost:22222 -s
> file*" and a connection should be forwarded through SSH to the waiting
> kermit process on the other end and all the "file*" files should be
> transferred.
> (I haven't yet figured out how to get SSH to do this -- ssh doesn't
> forward the port in the way I expected, but it should merely be a matter
> of getting all the command-line arguments right. Neither do I know if
> this is possible on non-Unix hosts, nor do I care, especially since this
> is just an academic exercise to prove it can be done -- nobody needs to
> do it because SSH already provides a more than adequate file transfer,
> and remote process connection, facility.)
Yes you could do it your way but that is much more complicated than
using Kermit as your local client (with ssh as the subprocess that
makes the connection) and then saying
kermit -s <file pattern>
kermit -g <file pattern>
to send or get files. Why should the user have to know anything about
tunnels?
Besides, ssh is supposed to be able to replace rlogin and I can use
Kermit to make a 'login' connection using the sub-process mechanism in
C-Kermit to start rlogin.
Jeffrey Altman * Sr.Software Designer * Kermit-95 for Win32 and OS/2
The Kermit Project * Columbia University
612 West 115th St #716 * New York, NY * 10025
http://www.kermit-project.org/k95.html * [EMAIL PROTECTED]
