[comp.protocols.tcp-ip] Telnet Commands in 3270 Data Stream

cavanaug@ncrcce.StPaul.NCR.COM (John David Cavanaugh) (04/23/88)

I wonder if anyone can help me with a protocol question.  Suppose you have
a Telnet connection that has been put into 3270 mode either by the 3270 Regime
option or a la tn3270 (Terminal Type, Binary, and End of Record options).
When you get a 3270 data stream, you look for IAC EOR to mark the end of it.
The question is:  What do you do when you find some other Telnet command
(e.g. IAC AO) before you find the EOR?  I can imagine four choices:

1.   Ignore the command (doesn't seem like a good idea).
2.   Handle the command immediately (the problem with this is that a lot
     of the Telnet commands don't have much meaning in 3270 mode).
3.   Save the command and do it after you find the EOR (doesn't seem right).
4.   Throw away the whole data stream and send an error message back to the
     user (which hoplessly messes up some poor data entry clerk's session).

My understanding is that Telnet commands "shouldn't" appear in the middle
of the data stream.  Is this correct?  Does any implementation of tn3270 or
the 3270 Regime option put commands there?

Thanks in advance for your help.  Reply to:

John Cavanaugh                          John.Cavanaugh@StPaul.NCR.COM
NCR Comten, Inc.
2700 Snelling Ave. N
St. Paul, MN  55113
   (612) 638-2822                       The opinions expressed ...

jbvb@VAX.FTP.COM (James Van Bokkelen) (04/26/88)

I can't say what is right, but I can say what we do:

In the incoming data stream, our client processes IAC regardless of IAC EOR
blocking (this also means we assume that IAC-stuffing applies within EOR
blocks).  The data stream we send to the server won't contain any telnet
command sequences within a 3270 block (after the AID but before the IAC EOR),
but it may contain IAC IAC.

You might want to pose this question on the TN3270 mailing list:

	tn3270@trantor.umd.edu	(I think it is trantor)

James VanBokkelen
FTP Software Inc.

davidc@TERMINUS.UMD.EDU (David R. Conrad) (04/26/88)

>You might want to pose this question on the TN3270 mailing list:

>	tn3270@trantor.umd.edu	(I think it is trantor)

Nope, other end... :-)

    tn3270@terminus.umd.edu

-drc