Open cdynak opened 8 years ago
That was a plan to separate to different functions in to separate ports. It is easy to imagine that user agent requests some action (for example push one button outside elevator) and in the same time controller sends command for some action to simulator. To provide supporting of simultaneous actions in different threads (listening and receiving) there is need to have two ports.
Changing that architecture right now is connected to joining two threads into one, building some kind of simple scheduler which will decide on what action port should be stopped in the idle state (listening or receiving). This is much effort to do such changes in that moment.
in my opinion using two different ports for communication is ambiguous when on one we only send and on second only receive. using only one port will also reduce some code.
in controller it is easy to change. what about simulator?