Bryon Welch writes:

>  I have used both telnet and NS/VT.  The major difference I have noticed
>  has been when making connections through the Internet.  NS/VT seems to hold
>  a connection better and if the connection is lost on the client end the
>  session on the HP3000 also exits.  With telnet I have ended up with a
number
>  of ghost sessions left on the 3000 when I lost the connection.

The "ghost session" problem that Byron writes about is/was real, but it has
been fixed. Indeed, it was corrected about a year ago within HP internally.
The bad news is that I don't know what release of MPE/iX that this fix will
first appear in. It may be in MPE/iX 6.0 now. If not, it will certainly appear
in MPE/iX 6.0 PP1.

Because we're a developer and because Jeff Bandle, the person at CSY's
Networking Lab is particularly accomodating, we and Adager and a few machines
at HP have been running off of several unnamed telnet patches for the last
year. Let me say that telnet is shaping up to be a real force. The "advanced
telnet" procedure that appears in QCTerm and which HP has very willingly
worked to implement into MPE makes an extraordinary difference for long-
distance telnetting. There is simply no comparison to "standard" telnetting.

As to reliability, because Adager kindly allows us unfettered access to their
machines, I've kept telnet sessions running on their machines for weeks on end
simply to test the reliability of continental-scale, long-distance telnetting.
The results have been for the last year, time after time, rock-solid
connections that make the Adager machines seem as if they sitting right next
to me.

All-in-all, I've become very enthusiastic about MPE's implementation of
telnet.

Wirt Atmar