gmail problem investigated: frame onload

[Allan Sandfeld Jensen]

On Tuesday 05 October 2004 21:39, David Faure wrote:
> On Monday 04 October 2004 18:53, Allan Sandfeld Jensen wrote:
> > I've been looking at how Safari handles it, and have come up with the
> > attached patch.
>
> Thanks for that patch - after discussion with Dirk, I based the final fix
> on it.
>
> Can you also check the code for unload? Did
> HTMLFrameSetElementImpl::detach() change, in particular? Thanks, it will
> save me the time to locate, download, and look into the latest webcore :)
<ggl:webcore>  ;)

No, it doesn't seem much has changed for unload. 
The only event differences in HTMLFrameSetElementImpl are the exclusion of getDocument()

@@ -470,23 +528,31 @@
             frameborder = false;
         break;
     case ATTR_ONLOAD:
-        getDocument()->setHTMLEventListener(EventImpl::LOAD_EVENT,
-	    getDocument()->createHTMLEventListener(attr->value().string(),"onload"));
+        setHTMLEventListener(EventImpl::LOAD_EVENT,
+	    getDocument()->createHTMLEventListener(attr->value().string(),"onload"));
         break;
     case ATTR_ONUNLOAD:
-        getDocument()->setHTMLEventListener(EventImpl::UNLOAD_EVENT,
-	    getDocument()->createHTMLEventListener(attr->value().string(),"onunload"));
+        setHTMLEventListener(EventImpl::UNLOAD_EVENT,
+	    getDocument()->createHTMLEventListener(attr->value().string(),"onunload"));
         break;
     default:
         HTMLElementImpl::parseAttribute(attr);
     }
 }
@@ -524,7 +584,7 @@
 {
     if(attached())
         // ### send the event when we actually get removed from the doc instead of here
-        getDocument()->dispatchHTMLEvent(EventImpl::UNLOAD_EVENT,false,false);
+        dispatchHTMLEvent(EventImpl::UNLOAD_EVENT,false,false);
 
     HTMLElementImpl::detach();
 }