Hi, Why do CDATA sections have seperate Start and End if their contents will always be treated as character data and reported by the CharacterDataHandler? What use could two handlers serve? Shouldn't the section have a single handler like the comments handler? Thanks