[eeps] Comment to EEP-24 Functions may be named using F/N in all module attributes

[Richard O'Keefe]

On 30 Oct 2008, at 4:55 am, Robert Virding wrote:
>
> Richards solution would probably be worse as it affects normalise/1  
> used anywhere.

For some reason I imagined that normalise/1 was local
to parse.yrl.  Of course it isn't.

I have some trouble understanding what else it could be
useful for, though.  The abstract forms that it accepts
are
	char, +char, -char,
	integer, +integer, -integer,
	float, +float, -float,
	atom, bin, string, nil, cons, tuple,
	record_field (?)

and for anything else (such as variables, function calls,
and control structures) it raises an exception.  So it
looked exactly like something that only made sense for
attributes, including types.