Two assumptions regarding the synthesis of communication and control policies for decentralized discrete-event control problems are revisited. In particular, the construction of information structures that yield certain communication strategies relies on a specific relationship between the languages describing the plant and the legal behavior as well as the respective automata for these languages. An example that satisfies part of the initial assumption but generates a invalid information structure is presented. Secondly, clarification of an assumption on the finite nature of some additional procedures used in generating communication and control policies is provided.