Greg writes:

> it seems obvious to me
>  that these phenomenon can be tested with a number of readily available
>  telnet clients, and some quite stupid telnet clients. The ones bundled with
>  various MS OSs, which probably do nothing whatsoever to accommodate or be
>  aware of MPE, and which probably use plain standard telnet. A server
problem
>  should show up in plenty of telnet clients, and a client problem should not
>  show up in any other readily available telnet client.

The problems that Alan is reporting are passwords being inappropriately
echoed and problems with block mode screens. Only the first condition (the
possibility of visible passwords) can be tested using the very simple telnet
clients that are supplied with Macs and PCs, but it does represent one more
test and it is certainly worth doing.

As it occurs, we have done that many times in the past with the Microsoft
telnet client, and again we have never seen any sort of problem with any of
the advanced telnet patches. There's just not that much that can go wrong.
Password echo is suppressed on the host, by the host program, using the
FCONTROL intrinsic, not by the client commanding it.

There were problems with the FCONTROL instrinsic and telnet host code several
years ago, but those problems have been corrected. The problems with telnet
that I've worked with Jeff Bandle on are these:

   o  Persistent ghost sessions due to abnormal telnet termination.
   o  Frequent timeouts when long-distance (continental scale) telnetting.
   o  The absence of XON/XOFF flow control.
   o  Eliminating the suppression of LF in CRLF pair in telnet.
   o  Double-echoing of typed text (the FCONTROL problem)

Jeff has been simply extraordinary in quickly addressing these problems and
correcting them, and he deserves an immense amount of credit. While I am
uncertain as to which of these features are in which patches, they're all
part of making "advanced telnet" work correctly and robustly.

Wirt

* To join/leave the list, search archives, change list settings, *
* etc., please visit http://raven.utc.edu/archives/hp3000-l.html *