Re: More over-engineering

[Paul Eggert]

Stefan Monnier wrote:
> The only advantage in this respect is that the representation of
> Lisp_Object is encapsulated inside Emacs core, so the same module can be
> used with a "emacs-32bit-narrow-int" and on "emacs-32bit-wide-int".
> The cost for that minor advantage seems way out of proportion.

I agree.  Although I'm a fan of --with-wide-int, I don't think it should force 
this extra complication in modules.  People who want Emacs with wide ints can 
build modules to suit.  This is the same as 32- vs 64-bit off_t, in traditional 
C libraries.