Request-Response

The Request-Response pattern can be used in several occasions: when dealing with server interactions like XHR calls, as a middleware between two applications, an internal communication loop between two independent processes etc.

It creates a very robust way of dealing with failures and tracking of events. By breaking down the communication into it's constituent parts and storing partial information on the state, it allows other processes to feed off this data in an elegant manner: progress bars, warnings, retry strategies, using the partial data received before completion, acknowledging to the user that the request was accepted, etc.

{
  requests: {
    items: {
      "xyz": {
        state: INITIALIZING | PROCESSING | RESPONDING,
        params: {},
        processing: {
          data: {},
          triggers: {},
          result: {},
          error: {}
        },
        requestedAt: timestamp,
        respondedAt: timestamp,
        error: {}
      }
    }
  }
}