Donna asks:

> does anyone know what the status is for wrq and
>  minisoft to support advanced telnet?        - d

I have no way of knowing Minisoft's or WRQ's plans, but let me explain
"advanced" telnet a bit, if you don't mind. There's nothing complicated about
it. It's merely a reinvocation of a 40-year-old idea: half-duplexed
communications.

When you strike a key on a keyboard using standard telnet, that character is
transmitted to the remote host, absorbed there, and then re-echoed back to
you. In both transmissions, coming and going, the packet tends to be a
single-character packet. Not only is this a little inefficient, it can be
slower than molasses, especially if the host is on the other side of the
world.

Under advanced telnet (and we chose to call it "advanced" rather than
anything with "half" in its name), when you strike a character on the
keyboard, it's still transmitted as a single-character packet, but it's also
immediately written onto your display by your terminal emulator. And when the
remote host receives the packet (1 to 1000 milliseconds later), it's absorbed
into the HP3000's input buffer, just as before, but it is NOT transmitted
back. Advanced telnet really is an advance, especially perceptually: only
half the traffic on the net is required now and you have instantaneous
response at your keyboard & screen.

Any telnet client can negotiate these conditions with its host if they wanted
to, thus WRQ and Minisoft can put this behavior into their code quite quickly
if they wanted to. However, there are two problems. One is due to a
double-echoing bug in the current HP3000 telnet code that virtually everyone
has (that has been fixed a year ago, but see below). And the other is that
passwords will now be visible if you have your emulator & host configured in
this necessary mode.

To correct the second problem, we put together a new draft RFC for telnet so
as to allow the transmission of a synchronizing signal from the host to the
client. It's called SLE (suppress local echo). When a password is to be
entered (and the HP3000 knows when that is; it's indicated by a FCONTROL
12/13 sequence), a DO SLE command is transmitted to the host. Once the
password has been entered, a DONT SLE is transmitted to turn the emulator's
local display back on. A copy of the draft is at:

    http://www.cisco.com/public/rfc/DRAFTS/draft-rfced-exp-atmar-00.txt

Although the draft has now expired, the SLE telnet option has been
permanently assigned. It will never go away and anyone who wishes to use it
can do so. It was our understanding with Jeff Bandle, the fellow at HP with
whom we worked, that everything was to be made available to everyone that
might want to use it. In that regard, I'm pleased to help provide as much
information as I can to anyone who might wish to implement the "advanced"
mode.


Ted Ashton asks:

> Is the patch which makes ^S and ^Q work general release yet?  Otherwise,
>  that's a significant problem.

I'm disappointed to say that I've recently found out that the telnet patch
we've had for a year now won't be released until sometime after 6.5. The
patch that we (and about a dozen other sites) have been using corrects the
following problems:

Two of the problems are in direct support of QCTerm:

     o corrects the FCONTROL 12/13 double-echoing problem
     o implements the SLE option necessary for password
        suppression under advanced telnet

The remainder are in general support of telnet on the HP3000:

     o institutes for the first time XON/XOFF flow control in telnet
     o corrects the LF suppression problem in a CRLF pair (this
        problem plaques Robelle's QEDIT; Reflection put a
        workaround in their WRQ code for QEDIT, Robelle later put
        a workaround in QEDIT for everybody else).
     o corrects the problem of persistent ghost sessions left by
        abnormally aborted telnet sessions
     o dramatically increases connection reliability over long-
        distance telnet connections

On the whole, this patch makes telnet enormously more usable on the HP3000
than the current release that is in all versions of MPE 5.5 pp4 to 6.0 pp1.
Until you receive this new version of telnet on the HP3000, you should
continue to use standard telnet mode in QCTerm, which is, of course, the only
version currently available in WRQ & Minisoft's emulators.

Wirt Atmar