Path: utzoo!utgpu!news-server.csri.toronto.edu!rutgers!usc!orion.oac.uci.edu!uci.edu!DHWalker
From: DHWalker@uci.edu (David Walker)
Newsgroups: comp.protocols.appletalk
Subject: Re: Parity in NCSA Telnet 2.3
Message-ID: <26743340.14849@orion.oac.uci.edu>
Date: 12 Jun 90 00:11:44 GMT
References: <21924@boulder.Colorado.EDU> <266D0F46.4DEE@intercon.com>
Organization: University of California, Irvine
Lines: 25

In article <266D0F46.4DEE@intercon.com> amanda@mermaid.intercon.com 
(Amanda Walker) writes:
> Parity is meaningless over a Telnet connection.  There may exist 
brain-damaged
> TCP/IP host implementations that try to do it anyway, but if so, they are
> broken.

Here's an application which I don't really think is brain-damaged (and why 
I would like to see the ability to set the high-order bit to emulate 
parity)...

We have a collection of dialout modems connected to a terminal server.  
People who want to make a call off-campus simply connect to that pool and 
give the AT commands for dialing.  We configure both the modems and 
terminal server ports for "eight bits, no parity" so that any parity 
required by the host at the receiving end of the call can be generated 
directly by the user.  This scheme works great as long as our user is 
connecting from a serial port on another terminal server or computer.  A 
telnet client that could emulate a serial line's parity bit would allow 
this to work for LAN-connected computers, too.

                               David Walker
                               Network Services Manager
                               UC Irvine
                                 DHWalker@uci.edu