From xemacs-m  Mon Mar 17 16:28:33 1997
Received: from CNRI.Reston.VA.US (CNRI.Reston.VA.US [132.151.1.1])
	by xemacs.org (8.8.5/8.8.5) with SMTP id QAA28861
	for <xemacs-beta@xemacs.org>; Mon, 17 Mar 1997 16:28:32 -0600 (CST)
Received: from newcnri.cnri.reston.va.us by CNRI.Reston.VA.US id aa23159;
          17 Mar 97 17:30 EST
Received: from anthem.CNRI.Reston.Va.US by newcnri.CNRI.Reston.Va.US (SMI-8.6/SMI-SVR4)
	id RAA11005; Mon, 17 Mar 1997 17:30:42 -0500
Received: by anthem.CNRI.Reston.Va.US (SMI-8.6/SMI-SVR4)
	id RAA06459; Mon, 17 Mar 1997 17:30:41 -0500
Date: Mon, 17 Mar 1997 17:30:41 -0500
Message-Id: <199703172230.RAA06459@anthem.CNRI.Reston.Va.US>
From: "Barry A. Warsaw" <bwarsaw@CNRI.Reston.VA.US>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
To: Mike Scheidler <c23mts@eng.delcoelect.com>
Cc: xemacs-beta@xemacs.org
Subject: Re: XEmacs needs a GUI toolkit interface. 
References: <9049.858636827@kocrsw12>
Reply-To: bwarsaw@python.org
X-Attribution: BAW
X-Oblique-Strategy: Allow an easement (an easement is the abandonment of a stricture)
X-Url: http://www.python.org/~bwarsaw


>>>>> "MS" == Mike Scheidler <c23mts@eng.delcoelect.com> writes:

    MS> I've never really dug very much into the nuts and bolts of
    MS> either package, but I'd imagine that the biggest hurdle would
    MS> be in setting up and managing the event loop.  Does anybody
    MS> have a feel for whether this is doable without making radical
    MS> changes?

What you could do with minimal pain (and all in elisp), would be to
define an interface at the lisp level, and run all the Tk goop in a
subprocess.  This has the advantage that you could cobble the thing
together with no changes to XEmacs at the C level.  I bet it would be
pretty useful too.  The disadvantage is that the abstraction might not
be right, and it would most likely be slower.  But you wouldn't have
pesky problems like merging event loops and such.

-Barry

