Tables discovered via process name, then window name, only going through two nested loops, instead of two nested loops followed by another loop.

(can someone test to make sure this actually doesn't break stuff on *nix?)

Windows HUD Stats Windows no longer appear in Windows taskbar (now gtk transients for table hud window)

pyfpdb/HUD_main.py

@@ -77,14 +77,14 @@ def process_new_hand(new_hand_id, db_name):
         hud_dict[table_name].update(new_hand_id, db_connection, config)
 #    otherwise create a new hud
     else:
-        table_windows = Tables.discover(config)
-        for t in table_windows.keys():
-            if table_windows[t].name == table_name:
-                hud_dict[table_name] = Hud.Hud(table_windows[t], max, poker_game, config, db_name)
-                hud_dict[table_name].create(new_hand_id, config)
-                hud_dict[table_name].update(new_hand_id, db_connection, config)
-                break
-#        print "table name \"%s\" not identified, no hud created" % (table_name)
+        tablewindow = Tables.discover_table_by_name(config, table_name)
+        if tablewindow == None:
+            sys.stderr.write("table name "+table_name+" not found\n")
+        else:
+            hud_dict[table_name] = Hud.Hud(tablewindow, max, poker_game, config, db_name)
+            hud_dict[table_name].create(new_hand_id, config)
+            hud_dict[table_name].update(new_hand_id, db_connection, config)
+        
     db_connection.close_connection()
     return(1)