Fix a bug that robotparser starves memory when the server responses
in HTTP/0.9 due to dissonance of httplib.LineAndFileWrapper and
urllib.addbase.
diff --git a/Misc/NEWS b/Misc/NEWS
index 9f7842e..7584d45 100644
--- a/Misc/NEWS
+++ b/Misc/NEWS
@@ -322,6 +322,8 @@
 Library
 -------
 
+- urllib.urlopen().readline() now handles HTTP/0.9 correctly.
+
 - refactored site.py into functions.  Also wrote regression tests for the
   module.