diff -r 82df0bb66ca7 -r 71ab68f31a1c sites/irclogs.qmsk.net/urltree.py
--- a/sites/irclogs.qmsk.net/urltree.py	Sat Feb 07 21:02:33 2009 +0200
+++ b/sites/irclogs.qmsk.net/urltree.py	Sat Feb 07 21:21:10 2009 +0200
@@ -263,7 +263,7 @@
         """
         
         # start with the defaults
-        kwargs = self.defaults()
+        kwargs = self.defaults.copy()
 
         # then add all the values
         for label_value in label_values :
@@ -403,7 +403,7 @@
         
         # found something?
         if not match :
-            raise URLError("No child found for label")
+            raise URLError("No child found for label: %s + %s + %s" % (self.get_url(), label, '/'.join(str(l) for l in label_path)))
 
         # ok, recurse into the match
         url, label_value = match.match(label_path)
@@ -414,7 +414,30 @@
 
         # return the match
         return url, label_value
-    
+
+    def get_url (self) :
+        """
+            Returns the URL for this node, by iterating over our parents
+        """
+        
+        # URL segments in reverse order
+        segments = ['']
+        
+        # start with ourself
+        node = self
+        
+        # iterate up to root
+        while node :
+            segments.append(str(node.label))
+
+            node = node.parent
+
+        # reverse
+        segments.reverse()
+
+        # return
+        return '/'.join(segments)
+
     def dump (self, indent=0) :
         """
             Returns a multi-line string representation of this Node
@@ -470,15 +493,13 @@
             Returns an (URL, [LabelValue]) tuple.
         """
 
-        # normalize the URL
-        url = os.path.normpath(url)
-
         # split it into labels
         path = url.split('/')
-
-        # ensure that it starts with a /
-        root_label = path.pop(0)
-        assert self.root.label.match(root_label), "URL must begin with root"
+        
+        # empty URL is empty
+        if url :
+            # ensure that it doesn't start with a /
+            assert not self.root.label.match(path[0]), "URL must not begin with root"
 
         # just match starting at root
         return self.root.match(path)
@@ -488,10 +509,13 @@
             Looks up the request's URL, and invokes its handler
         """
         
-        # get the request's URL path
-        url, label_values = self.match(request.get_page_name())
+        # get the requested URL
+        request_url = request.get_page_name()
+
+        # find the URL+values to use
+        url, label_values = self.match(request_url)
 
         # let the URL handle it
-        url.execute(request, label_values)
+        return url.execute(request, label_values)