Re: 3270 telnet option?

Bob Braden (
Sun, 17 Aug 86 15:11:44 PDT


                I have heard rumors of a 3270 emulation package
        intended for use as a telnet option. Does such an animal
        exist? If so, who wrote it? Is it available? Is there
        published literature describing it? The rumors I heard
        associated it with the UCLA IP implementation for MVS, if
        that is any help. This may be rather garbled, to say the
        least; my apologies for the lack of specifics.
        Mike Murphy GTE Laboratories, Waltham, MA


There is an accepted mechanism for supporting full-screen IBM 3278 operation
over a Telnet connection. The essential idea is that you ship the 3278
order stream transparently, after negotiating the Binary Transmission
option in both directions. There is an initial Terminal Type negotiation
to establish the model number, and both ends must agree that they handle
End-of-Record marks (which are used to delimit keyboard and screen

This mechanism is currently supported by the User and Server Telnets of both
the Wisconsin VM code and the UCLA MVS code. Unforunately, there are
minor differences which make the two versions almost but not quite
compatible. No one has "bit the bullet" to reach a standard and document it.

In addition, there are programs which run under Unix to act as the user side
of the full-screen Telnet interaction, emulating a 3278 on the local ASCII
CRT/workstation. The one I am familiar with is called tn3278, and was
written by Greg Minshall at Berkeley. It is available by public FTP from
Berkeley, although you should be aware it needs several bug fixes.
I recently installed it on my Sun 3, and after I fixed the most important
bugs, it provided excellent service to the UCLA 3090 running MVS (when
you have to communicate through a loaded ARPANET/MILNET gateway, full-
screen interaction works A LOT BETTER than character-by-character mode!!!).

Bob Braden

This archive was generated by hypermail 2.0b3 on Thu Mar 09 2000 - 14:36:35 GMT