Re: (setf capi:pinboard-pane-position) doesn't change position?

I wish I had an unsend button—it looks like making sure that
everything happens in the pane process (i.e., using
capi:apply-in-pane-process) fixes the problem.

On Fri, Mar 12, 2010 at 2:22 PM, Joshua TAYLOR <tayloj@cs.rpi.edu> wrote:
> I've written a description of a pinboard layout to file and while
> trying to recreate the structure, I occasionally end up with objects
> at 0,0 rather than the position they're supposed to be at.  I haven't
> been able to track down the exact cause yet, and it's not reliably
> reproducible.  As such I'll try to follow up when I have more
> information, but just in case this is something someone's seen before,
>  any ideas what might make something like this fail?
>
>                   (setf (capi:pinboard-pane-position vertex)
>                         (values x y))
>                   (multiple-value-bind (xx yy)
>                       (capi:pinboard-pane-position vertex)
>                     (assert (and (eql x xx) (eql y yy))))
>
> I.e., immediately after setting the capi:pinboard-pane-position of a
> vertex, the values returned by (capi:pinboard-pane-position vertex)
> don't match.  Invalidating the pane constraints between setting and
> retrieving the values makes the assertion fail less often, but doesn't
> completely prevent it.
>
> Two things that might be relevant:  the pane isn't displayed on screen
> while this is happening, and the position might be outside of the
> current boundaries of the pane, making it grow.
>
> Thoughts?  Similar issues?
>
> Thanks in advance,  //JT
>
> --
> Joshua Taylor, http://www.cs.rpi.edu/~tayloj/
>



-- 
Joshua Taylor, http://www.cs.rpi.edu/~tayloj/