[cells-gtk-devel] Re: Cells-gtk now UFFI-ized, running under AllegroCL and Lispworks on win32, next stop OS/X.

[Kenny Tilton]

Andras Simon wrote:

> The following hack almost solves the tree demo slowness problem.
>
> diff -b -u -r /tmp/root/gtk-ffi/gtk-core.lisp ../gtk-ffi/gtk-core.lisp
> --- /tmp/root/gtk-ffi/gtk-core.lisp     2004-12-16 05:51:17.000000000 
> +0100
> +++ ../gtk-ffi/gtk-core.lisp    2004-12-24 01:02:20.444497200 +0100
> @@ -67,8 +67,15 @@
>  (defmacro with-g-value ((var) &body body)
>    `(call-with-g-value (lambda (,var) , at body)))
>
> +#+cmu
> +(ffx:def-type g-value-type
> +             (* (alien:struct gtk-ffi::g-value
> +                              (gtk-ffi::g-type (array (alien:signed 
> 32) 16)))))
> +
>  (defun call-with-g-value (fn)
> +  (declare (optimize (speed 3) (safety 0) (space 0)))
>    (let ((gva (ffx:fgn-alloc 'g-value 1 :with-g-value)))
> +    #+cmu (declare (type g-value-type gva))
>      (unwind-protect
>          (progn
>            (dotimes (n 16)

Cool, thx. I will work that in.

>
> Can someone translate the above deftype into a more uffi-friendly 
> form, so it doesn't have to be conditionalized?
>
> Kenny, I see that you didn't #-cmu-ized out the tree demo in 
> test-gtk.lisp. I wonder if people can sit through the minutes (hours?) 
> it takes the demo window to come up... 

uh-oh. I must have just missed it. Was that in a patch, or was I being 
given undue credit for ability to handle the obvious on my own? :) 
Probably the latter. Well, do we still need it disabled after the above? 
If so, you are talking about the whole "tree-view" tab, yes? Not just 
the treebox sub-tab?

kt

-- 
Cells? Cello? Celtik?: http://www.common-lisp.net/project/cells/
Why Lisp? http://alu.cliki.net/RtL%20Highlight%20Film