Should redispatching from within Reducers be allowed to prevent the double up of checking state with the middleware and then reacting to it with a redispatch?
In the current model, you're both reacting in a reducer to the state change, and preempting that state change in middleware before it even happens. This is far from ideal.
Should redispatching from within Reducers be allowed to prevent the double up of checking state with the middleware and then reacting to it with a redispatch?
In the current model, you're both reacting in a reducer to the state change, and preempting that state change in middleware before it even happens. This is far from ideal.