Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!samsung!olivea!oliveb!bunker!wtm
From: rbarth@tumtum.cs.umd.edu (Dick Barth)
Newsgroups: misc.handicap
Subject: Re: TDD bulletin boards?
Message-ID: <14798@bunker.UUCP>
Date: 9 Oct 90 02:00:53 GMT
Sender: wtm@bunker.UUCP
Reply-To: rbarth@tumtum.cs.umd.edu (Dick Barth)
Distribution: misc
Lines: 37
Approved: wtm@bunker.UUCP

Index Number: 10947

Mark Becker (mbeck@rice-chex.ai.mit.edu) writes:
   In article <14620@bunker.UUCP> rbarth@tumtum.cs.umd.edu (Dick Barth) writes:
   >Index Number: 10787
   >
   > One well-known brand (which shall remain nameless) mis-translates
   > an ASCII null into a backspace character.

   Dick, I thought this was a *feature* of the non-mechanical TDD's.
   Don't all the electronic ones do this?

Lord, I hope not. It's WRONG, WRONG, WRONG! 

   If not.. yeesh.. oh well.  The BBS software I modified, as far as I
   know, doesn't output that code at all when dealing with TDD/Baudot
   callers.

   Regards,
   Mark Becker
   mbeck@ai.mit.edu

Mine doesn't output nulls either in Baudot mode. It accepts Baudot nulls
from TDD callers and treats them as a backspace, but of course you don't
echo the incoming from a TDD caller and so a null is never SENT in Baudot.

All this raises the question of what do you do when you actually want to
send a null to a TDD. They're really needed when talking to a mechanical
teleprinter, since those things can't actually do a CR/LF in the space of
two character times. I remember from my Signal Corps days that when 
punching tape for one of those clunkers you had to puch two CRs, a LF 
and a couple of LETTERS characters to kill time while the carriage slid
back.

What I use for a NULL is actually an upshift or a downshift, depending
on what case the TDD is actually in at the moment. It kills time and
changes nothing, so it serves the purpose of a null very well.