I finished the prototype for CrawlerDaemon, https://github.com/Cyan3/oodt/commit/a54f05c9111ee8d2bc95e1831a1bcc8fe7073268
I have a question, for now on I can take tow paths: - Implement avro-rpc through aggregation, like in catalog module, where AbstractCommunicationChannelServer/Client separate application logic from server logic, and then extend and implement avro logic. - Implement avro-rpc where application logic and server logic be separated. Which approach it is better?
