Hmm.  It now appears that even the demo code won't run after being run
from the Editor once.  I swear it did in the past.

The bindings for *standard-input* and *standard-output* are indeed
different depending on where the form is evaluated, but they are both
instances of the same class:RUBBER-STREAM.  I guess it makes sense
that they would be different since they're going to different
output-panes.

If I wait around a couple of minutes, the Gtk window will come up, but
will refuse to close and GNOME will tell me that it's not responding.

I haven't had a lot of time to look at this the last few days, but
maybe tonight I'll rummage through the innards of cells-gtk and try to
figure out what's going on.  This is too weird to just let it go.

(BTW, cells and cells-gtk are pure glory.  Thanks, Ken and Vasilis).

On 8/14/06, Ken Tilton <[EMAIL PROTECTED]> wrote:



On 8/14/06, Bill Atkins <[EMAIL PROTECTED]> wrote:
> Nope, I'm running from the LispWorks IDE.  I'm getting by with SBCL
> and SLIME at the moment, but I'd prefer to be back in LW.  Not sure
> what else to try.


Wild guess: Have you looked at the runtime bindings of various special
variables  dedicated to input/output? Working with a different Lisp
altogether and having similar REPL-OK, other-not issues, the folks at tech
support ended up explaining that the IDE did cute things in re stuff like
*standard-input*.

But I am especially fascinated that the gtk-demo does not have this problem.
I guess whatever difference the IDE creates running from there (such as the
thread thing or some special binding) is not touched by gtk-demo code but is
by yours. But did you call up all the different tabs?

I would look especially at any difference in how the demo and my windows
closed (I mean, the code that runs at close time) in the event that the
issue is how gracefully one exits.

hth, kt





--
Bill Atkins
_______________________________________________
cells-gtk-devel site list
[email protected]
http://common-lisp.net/mailman/listinfo/cells-gtk-devel

Reply via email to