CLIM mail archive
Re: Experiences with CLIM
Date: Thu, 8 Nov 1990 10:13-0500
From: Scott McKay <SWM@sapsucker.scrc.symbolics.com>
Subject: Experiences with CLIM
To: firstname.lastname@example.org, clim@BBN.COM
Date: Thu, 8 Nov 1990 06:14 EST
From: Bernd Wild <email@example.com>
Ok, we have received CLIM 0.9 from Franz for ACL 3.1 about 20 days
ago. Since there we have done some testing with it to compare it against
the CLIM prerelease on Symbolics.
Which CLIM prerelease on Symbolics, just out of curiosity?
First, CLIM is relatively slow, i.e.
if you push your frame via X on an X-Server CLIM blocks that server
for a few seconds before it starts to paint the frame.
Lest CLIM undeservedly acquire a reputation for slowness, I should point
out the Symbolics CLIM running on Genera beats the pants off of Dynamic
Windows in terms of performance. Symbolics CLIM running on 386's under
CLOE also has surprisingly good performance, comparable to Dynamic
Windows running on my 3645. Furthermore, in addition to improving the
general robustness of CLIM, we have been steadily improving the
performance, and anticipate that when we more broadly release CLIM
with Genera 8.1 (and the next minor release of CLOE 3), it will be
Perhaps one difference is that CLIM has not been tweeked for
performance on stock hardware LISPs like Franz or Lucid to the extend
that it has for Symbolics machines. From my experince with CLIM in Lucid,
there is a lot of room for improvement here.
We succeeded in
starting the address book demo and the CAD demo, but when resizing the
x-window no refresh happened although the mouse-sensitive regions
were still present on the now empty frame.
I hope that these bugs will be removed in the next version
CLIM is evolving from prototype to product, and fixing reams of bugs is
a natural part of that process.
Perhaps these bug fixes should be shared over this list.
understand the comment ILA is making not to use 0.9 for commercial
Main Index |