Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!mnetor!uunet!seismo!mcvax!steven
From: steven@cwi.nl (Steven Pemberton)
Newsgroups: comp.emacs
Subject: Re: gnumacs bindings
Message-ID: <69@piring.cwi.nl>
Date: Tue, 22-Sep-87 07:55:39 EDT
Article-I.D.: piring.69
Posted: Tue Sep 22 07:55:39 1987
Date-Received: Thu, 24-Sep-87 03:01:23 EDT
References: <3590005@hpindda.HP.COM> <8709181153.AA08300@ucbvax.Berkeley.EDU> <636@unmvax.unm.edu>
Organization: CWI, Amsterdam
Lines: 59

In article <636@unmvax.unm.edu> mike@turing.unm.edu (Michael I. Bushnell) yells:
> Go to your login prompt.  GUESS WHAT???? DEL works!!!!!! So does ^H!!!!!
> THEY BOTH WORK. UGH!!!!!
> There is NO predisposition in UNIX for BS instead of DEL!!!!
[...]
> But Emacs is strapped for keys.  So it has to use only one key per
> function, and the most logical choice is DEL.  
Not true. You can use two keys for the same function with no trouble.

> Look at the ASCII specification!!!!  It quite plainly states that
> DEL means "ignore last character,"
The ISO specification that I have in front of me says thet DEL is used
to erase or obliterate an erroneous or unwanted character in punched
tape. This would suggest (if we were trying to retain compatibility
with the original definitions of these characters) that delete-char
would be a better binding for DEL. But we ain't, so we can choose to
use them how we like.

> ACK!  UNIX could care less what you use.  Really.  It is quite up to
> you.  Fortunately, EMACS lets you pick.

The Unix behaviour, I contend, is the better. There are two equally
used standard keys, and Emacs should accept both of them *as default*
if we want to encourage new users to try emacs.

Here where I work, there are a dozen departments. Three of these are
computer departments where most of the members are competent enough to
deal with bindings and the like. However, the other nine departments are
non-computer people who do however do a lot of editing. I think the
^H=help binding is the biggest obstacle to them using gnumacs. If they
should go and try emacs, and then find that they can't use backspace
(which IS the default stty setting here), they are not going to bother
to find out how to rebind. They're just going to go back and use
whatever they're used to, declaring emacs to be "unusable". I'm sorry,
but that is just how it is. I really do know people who do not use
gnumacs because backspace means help; they use computers for their
work, and they can't be bothered to find out how to rebind keys just
in order to see if they want to use emacs.

I might say that ^S also causes a problem in this respect. We have
terminals here that send ^S/^Q. If we went to management and said "we
want new terminals, so we can use emacs", they will say "use a
different editor".  Besides, people like the terminals we have, and
couldn't care less whether they send ^S/^Q or not.

The point I'm trying to make is this: the default bindings affect the
user acceptance of Gnumacs. We have to face up to the fact that
potential users are being frightened away from using it simply because
the default bindings do not encourage them to try it.

I might also remark that I only started using gnumacs because I was
forced to: the emacs we used to use didn't compile on our new machine.
It was one hell of a job trying to work out from the manual how you
changed the binding of ^H to backspace. It is quite clear that ^H is a
prefix key bound to a keymap, but how to bind another key or
key-combination to that keymap? I would like to suggest that something
be said under "customization" in some future version of the manual.

Steven Pemberton, CWI, Amsterdam; steven@cwi.nl