Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!mnetor!uunet!husc6!cmcl2!rutgers!ucla-cs!zen!ucbvax!MCMASTER.BITNET!BEAME
From: BEAME@MCMASTER.BITNET
Newsgroups: comp.protocols.tcp-ip
Subject: Re: TELNET <CR> <LF>
Message-ID: <8708110359.AA17570@ucbvax.Berkeley.EDU>
Date: Mon, 10-Aug-87 22:58:00 EDT
Article-I.D.: ucbvax.8708110359.AA17570
Posted: Mon Aug 10 22:58:00 1987
Date-Received: Wed, 12-Aug-87 01:58:10 EDT
Sender: daemon@ucbvax.BERKELEY.EDU
Distribution: world
Organization: The ARPA Internet
Lines: 21


   When I was converting my PC VT100 emulator to run TELNET TCP/IP, I found
myself with a problem. My emulator allows the user to define ANY key to
mean ANYTHING. Therefore the big key with the arrow ( <-- ) on it did not
always mean the End-Of-Line character. What should I send when it was hit ?

   Though it violated the spec. I decided not to have an EOL key, but to send
exactly what the user requested when he defined the keyboard. The normal
mapping had <CR> go to <CR> and <LF> go to <LF>, no conversion was done.

GUESS WHAT !!!! When running to BSD 4.2/4.3, TWG, Ultrix 1.2, KNET and
Bridge CS/100's to VMS vaxen and an IBM 7171, EVERYTHING worked. The VMS
vax recognised the <CR> as a <RETURN> on a terminal and the <LF> as
delete the last word. BSD and Ultrix had no problems in determining what
was meant by the user.

Could this not be an interm solution till all is well (assuming all is not
considered well now) ?

     Carl Beame
     BEAME@MCMASTER.BITNET