XML KAction problem.

Zack Rusin zack at kde.org
Wed Sep 25 03:23:41 BST 2002


On Tuesday 24 September 2002 18:51, Simon Hausmann wrote:
> Hm, strange, very strange. Can you try if reverting my patch changes
> anything?

Does 'doesn't crash anymore' count as a change? ;) I reverted and 
checked what's wrong. Here's something funny - your patch fixes the 
dangling pointer problem by going through the guarder pointer but even 
after deletion of the area a call to removeDockWindow is a requirement. 
Beats me what happens but the following change in the destructor of 
QDockWindow fixes applications:
-    if ( area )
-       area->removeDockWindow( this, FALSE, FALSE );
+    if ( this->area() )
+       this->area()->removeDockWindow( this, FALSE, FALSE );
it's obviously wrong because we go back to dealing with dangling 
pointers and I sure as hell am not sure why the call to 
removeDockWindow is a requirement considering that the area shouldn't 
even exist. Weird...

> :-( (but maybe you could remind me then, preferably on IRC? :)

Sure, I'll try to spend some more time on it myself.

Zack

-- 
Why is abbreviation such a long word?





More information about the kde-core-devel mailing list