Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!watmath!clyde!burl!ulysses!harpo!decvax!mcnc!ecsvax!taylor
From: taylor@ecsvax.UUCP
Newsgroups: net.dcom,ecsnet.general
Subject: looking for the perfect escape sequence
Message-ID: <2342@ecsvax.UUCP>
Date: Thu, 19-Apr-84 11:56:35 EST
Article-I.D.: ecsvax.2342
Posted: Thu Apr 19 11:56:35 1984
Date-Received: Fri, 20-Apr-84 02:19:10 EST
Lines: 51


Some new switching equipment which we are installing has a feature by which
the user can enter a two-character ascii sequence to

1) disconnect from the current session, or

2) temporarily disconnect in order to turn echoplex on or off.

The sequence is fully definable, and the only restriction is that the
first character of each of the sequences must be unique.  In order that
the character be reproduced in the same fashion on every terminal, the
character should probably be a control character for one of the
characters 40 through 5f.  (This includes the entire 26 character
alphabet and a few other characters.)

As soon as the first character of the sequence is encountered, it is buffered
until the next character is entered.  If the next character is the second
character of the sequence, the appropriate action is taken.  If the second
character is not the second character of the defined sequence, both
characters are then passed on to the host.

The problem is....

HOW DO WE FIND A 2-CHARACTER SEQUENCE THAT NOBODY ELSE USES?

For instance, EMACS uses all characters in this range except cntl-\.  Escape
is also a possibility as the first character, since it is virtually always
used in a two-character sequence, but the problem then becomes, what should
the second character of the sequence be?

At this time, the leading contenders for the world's most innocuous
sequences (which could be remembered by users) are:

Sequence 1-	cntl-\	cntl-\

Sequence 2-	escape        carriage return

sequence 3-	escape        cntl-\

Has anyone else faced this problem?  What were your results?

p.s.  yes, we can disable this function on a terminal-by terminal basis,
but we are striving for as much uniformity as possible in a large network.

Thanks!

Steve Taylor
NC Educational Computing Service
!mcnc!ecsvax!taylor

(919)-549-0671