I agree that states are probably a much better option than viewstacks because they are more dynamic. I probably will consider doing this as I move along for this reason. I guess I was just wondering if there was a better way than using constants in the modellocator to bind what state, or selected index, that a component is in. If I used states, am I correct in saying I would still need to have a static variable in the modellocator that each views state variable would have to bind to? I am thinking there must be a better way to decouple the view from the model in this instance. Thanks, I appreciate this great feedback.
--- In [email protected], "Justin J. Moses" <[EMAIL PROTECTED]> wrote: > > For your application, have you considered separating your views within > states, and only populate them on the "EnterState" event? You could > employ modules that load up on that event, rather than preloading them > with components. > > The advantage of using states is that when your user comes back to > opened states, they show themselves as they were left. Although the > obvious caveat is the memory they take up (which you'd need to analyse > in the Flex Profiler - assuming ur using Flex Builder 3 Pro). > > You can create the states dynamically, say when your user logs in, to > load up the modules that type of user might need. >
