Re: Upcoming loss of usability of Emacs source files and Emacs.

[Paul Eggert]

Richard Stallman wrote:
> I think we should aim to represent Lisp code as ASCII so it can be
> edited without an input method.

That makes sense for Lisp code itself, but not so much for strings and comments 
that Lisp code contains.  Years ago we started putting non-ASCII characters into 
Lisp strings and comments, and this has worked well.  Going back and imposing an 
ASCII goal or requirement now would would be a maintenance pain for little 
benefit.  For example, it would be much harder to read and change this 
ASCII-only Lisp code:

  (string-match "[^a-zA-Z\u03B1-\u03C9\u0391-\u03A90-9']" name)

than to read and change what's in calc-explain-units-rec now:

  (string-match "[^a-zA-Zα-ωΑ-Ω0-9']" name)

Of course the fact that non-ASCII characters work in Lisp strings does not mean 
that we should necessarily change our source-code strings.  My point here is 
only that the goal of ASCII-only Lisp source files is now often more trouble 
than it's worth, and it should not be a determining factor when considering what 
source-code strings should look like.