Re: Proposal: Bind backspace and delete like DEL in view-mode

[Kim F. Storm]

Richard Stallman <address@hidden> writes:

> The change would effectively eliminate the feature, which we worked so
> hard to develop, of adapting Emacs behavior to the specific keyboard.
> That would be a big step backwards--no thanks.

My major concern is that if you have made great efforts to make everything
work smoothly  (by intelligently mapping <backspace> and <delete> in the
be possible way),  then it shouldn't be so easy for the average user to
break that. 

Since it is currently much easier to express the syntax which is
necessary to remap backspace and delete - than DEL - users (like I
did) will do just that - and thus break the modes which only binds DEL
(as they should do).

address@hidden (Kai Großjohann) writes:

> Right now, Emacs translates from <backspace> to DEL on some keyboards.
> I was proposing to translate from DEL to <backspace>, instead.  I
> think that this does not remove features.  (Just to clarify.)
> 
> But it would be a big change, since a lot of keymaps have to be
> changed to use <backspace> rather than DEL, and I guess that you
> wouldn't approve of this.  After all, the current mechanism works
> quite well.
> 
> I hope that Kim can live with this :-)

It wasn't me who suggested to reverse everything :-)

I see two ways to improve (or fix) the currrent behaviour:

1) Modify define-key so that it will recognize DEL, TAB, etc as
   synonymous for the actual characters.

   This is fully backwards compatible, and it will allow users to
   use the same key names as emacs reports for the corresponding
   keys (e.g. DEL is reported as the key bound by C-h k backspace).

2) Modify define-key so that, when binding a key which maps to DEL in
   function-key-map, DEL is mapped instead - unless a new optional
   argument NO-REMAP is non-nil.

   This is not truely backwards compatible, but from a user's point of
   view it will work as before (except that it doesn't break the
   use of the backspace key as DEL in the mode keymaps).

I'll look into implementing both of these proposals so we can see
whether it is feasible.