[Top][All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: tools for GNU Emacs as application framework

From: Tom Lord
Subject: Re: tools for GNU Emacs as application framework
Date: Thu, 25 Jul 2002 11:41:54 -0700 (PDT)

   This seems like a neat hack.  It raises two issues:

   1. Emacs contains a widget system, widget.el.  It looks like Etalk is
   cleaner than the mechanisms used for widget.el, so perhaps it would be
   good to replace all that with your code.  However, your widget system
   needs to have all the facilities of the existing one.  Then, part of
   installing your widget system would be making cus-edit.el use your
   widget system.

Right.  And since our thread is interrupting g-e-s, let me at least 
point out to readers there that I am not personally going to get that
particular task anytime soon -- so it is up for grabs.

By the way, I only tried `M-x customize' recently for the first time.
It's so nice.  If you look at the old Monkey code, you can see it goes
off on a complete tangent relative to that design (and other recent
parts of Emacs) and is clearly not ready to even think about putting
in ./lisp (e.g., "modal.el" is a bad idea even though it aims for some
useful functionality).  That's part of where Etalk comes from - a new
foundation to do monkeyish things that works better with the way Emacs
works these days.

   2. People are doing some real work towards rehosting Emacs Lisp on

Yez.. well.  Is this part of Guile?

I've done a lot of thinking over the past few months (well, years, but
especially over the past few months) about this design space.  There's
some things I think should be going differently than they are, but I
don't think any of the relevent developers are much interested in
hearing about it or in deviating from their ongoing work (and who can
blame them?).

So, this is an awkward area.

   If we do that, how would it affect Etalk?  Would it be
   possible to define classes in Scheme and define subclasses using Lisp,
   for instance?

Uh...what are you really asking here?   Sure.  Etalk is really simple
and based on simple representations -- it should be doable in just
about any lisp-ish environment.

If you mean: "is there a really clean way to put all these pieces
together?"  I'd bet there is.  We could talk about my own Scheme plans
if you like.   Does anyone else really care about this issue?  I'm not
so sure, sometimes :-)


reply via email to

[Prev in Thread] Current Thread [Next in Thread]