I'm trying to implement a popup window (NativeWindowType.UTILITY) in an AIR 2.7 application that uses Swiz for dependency injection.
I have followed the Swiz guidelines that I've been able to find, and implemented ISwizAware on the class that creates the window, and I am calling _swiz.registerWindow() before opening the window, and dependency injection works fine on the window itself after this.
However, the problem I am running into is that I have a child view within that window, and I have a mediator that uses the [ViewAdded] and [ViewRemoved] tags. Neither the view added nor view removed functions are triggering. I'm thinking the issue is either:
The child view is not correctly registering with Swiz.
The swiz instance doesn't know about the beans (I have tried manually adding the bean however, which didn't have any effect).
The ViewAdded and ViewRemoved metadata tags simply aren't working because each NativeWindow object has its own stage instance.
Anyone know more about this?
Popups are a special case since they don't get added under the same display tree as your application. Under Stage (the main wrapper for Flash Player), you'll have Application where your code resides for Swiz, but Popup is in a separate layer above Application. Since they're siblings, Swiz cannot listen in for when the popup is being added to the Stage.
The way around this is to either set the properties of the popup manually (which is normally the easiest way) or manually add the popup to Swiz's awareness. For this you'll have to look at the documentation since I haven't touched Swiz in a long time.
Related
ok so I have a TitleWindow that I open up... and I have 6 states defined. I am using the Presentation model pattern for all of my views. I have discovered a frustrating nuance. When I tell my window to go to XXX state, the controls have to initialize since the states in flex use lazy loading. so... my PM code that says myTextArea.text bombs out and say "cannot access..." so as a work around, I made some creationComplete events on my controls to register the control with the PM. So when the state changes, the textarea finally initializes and on creationComplete calls PM.registerTextArea() which sets a reference to it. then in that function I run my code... myTextArea.text.. etc.
This seems like such a ugly hack that I hate it. Is there any way to wait until the entire state in created then call code on the PM? I have tried enterstate... but this event seems to fire before the state controls are ready.
I tried to add a comment, but I guess editing is the only way I can do this...
to everyone: thanks for the great feedback. I am doing something slightly off straight PM. Every view has a viewController (as I call them). Its kinda my own hybrid of a delegate / dataprovider. But this is moot. It's the flex component lifecycle when dealing with states that is the pain. if you change state.. there is no event to say "all my components in this state are ready". only event to say "we changed state". so my code that fires on state change is trying to talk to components that aren't ready yet. So from what I see, there seems to be no design pattern or perfect way to ensure that all components are created in a state unless using creationComplete on every component in the state to register it is ready... if you don't, you get a race condition. Regardless of frameworks or design patterns, this seems to be a root issue.
The textarea is an easy PM fix.. just bind it to the pm value. But there are other times I can't.
Specifically, I am trying to attach video to a display once I get to that state. This is done via addchild. regardless of where I do it.. I need to know that the videoDisplay is done loading before I call addchild. I even tried currentStateChange event since docs say that fires last... but alas.. the components in the state are still initializing. So it seems that creationComplete is my only option. Maybe the only sane way to keep to clean code is to create the entire thing (videodisplay and video) using as once the state is entered. I just hoped the flex framework had events to ehlp me here rather than buiilding everything on the fly in as.
Since your PM has a reference to a visual component (myTextArea), this isn't a completely pure form of a presentation model. It appears to be more of a supervising presenter / controller type of setup.
That being said, the way I would fix your problem would be to use a complete presenter outright ( no knowlege of the view at all ) or use a complete controller ( view implements an interface through which the controller communicates ). The advantage of using a presenter in Flex is that you can create a bindable value such as text or selectedItem, and the view will bind to that variable whenever it comes online so the issues dealing with the lifecycle of Flex components go away.
I'm moving some Flex classes to pure Flash/AS3 to be part of a shared library. I have an event listener on FlexEvent.CREATION_COMPLETE to create and add display-objects once things are running, but I don't know the right way to do this in Flash?
edit: Maybe if my class subclasses Sprite rather than a Flex class, I can add children in the constructor, I don't need to wait for a "construction complete" event?
It depends on what you are trying to do: When you add children in the constructor, you will not be able to depend on any information about the stage and/or other elements on it. I have found it a better practice to add an event listener for Event.ADDED_TO_STAGE in the constructor, and have the handler method initialize the element.
You're really comparing apples and oranges here.
The Flex Framework is a UI Framework, and the CREATION_COMPLETE event is part of that framework.
Flash Professional is an IDE, primarily geared for creating animation. It has ActionScript support, but is not a framework and does not offer anything beyond native ActionScript other than libraries or classes you build, buy, create, or borrow.
If you want some parallel to creationComplete, you'll probably need to find--or create--a UI Framework. It is, in theory, possible (although not easy, nor something I'd recommend) to initialize the Flex Framework for use inside a Flash IDE created movie. After all, under the hood it is all ActionScript.
creationComplete Dispatched when the component, and all of its child components, and all of their children, and so on have been created, laid out, and are visible.
See more info on the flex life-cycle events and some that are plain flash here: also +1ed the added_to_stage as it's likely what you'll need in this case.
http://livedocs.adobe.com/flex/3/html/help.html?content=containers_intro_3.html
In the examples for RobotLegs, it appears that mediators are used on every button/textArea, rather than on the custom component that contains these children. This would be very time consuming would it not?
From Joel Hooks InsideRia Example
Dependency injection works better with
unambiguous classes. What this means
is that by extending TextArea into our
new MessageView class, we are creating
a specific view component for the
dependency injection to act upon. This
is important if our application were
to have several TextAreas that served
different purposes. By dividing up our
classes in this way, we are clearly
defining the intent of the class and
allowing for the dependency injection
tools to do their jobs effectively.
No, don't mediate every child component. Your components should be organized into groups that perform related actions. In the examples the components are extremely simple and do not reflect what a real application would look like.
One rule of thumbs I use is thinking if that component needs any communication with the rest of the application, or if its only a part of a whole. Keep in mind that mediator are only intended to serve as a bridge between the view and the app.
For example, if I've a view with a form (asume a login form) I don't mediate all the child components (the textfields, the buttons, etc.) because it would be pointless and would have a proliferation of classes and objects on runtime. When I do the form I think, what does the view by its own? and what the other parts of the app should do with it?
When the user fills the form and clicks a button, the view dispatches an event (LoginRequestEvent, for this case), and then the mediator should redispatch that event, making the mediator very lean.
But with practice of the framework, you'll come up with this feel of what you shold mediate. For instance, in one app I mediate every item renderer of a list, and on other I mediate a view stack with two or three navigation contents.
Hope it helps
When I try to access the hidden TABs of my tab navigator control in action script, it returns a null error. But it works OK if I just activate the control in the user interface once. Obviously the control is not created until I use it. How do I make all the tabs automatically created by default ?
<mx:TabNavigator creationPolicy="all"/>
That should do it. Deferred instanciation is a feature, but sometimes it is a hassle.
The Flex framework is optimizing creation be default (creationPolicy="auto") so if you have a configuration dialog with a lot of tabs, for example, and the most useful tab is the first one, your application does not spend time and memory initializing the tabs that the user never sees.
This makes a lot of difference when dialogs like this never release, and is a good default to go with.
One thing to look at is using a private variable in your dialog/form instead of pushing the data to the control on the hidden page. This style treats the whole form as if it were a component, which it sort of is. To repeat: the MXML form/dialog/canvas is a class, and it can have data and methods in addition to containing other components.
Cheers
On a side note, I've run into the deferred-loading policy in a multi-state application, and circumvented it by forcing all elements to be included and invisible in the initial state. Something to consider, but only as a hack.
I am having a problem with preloaders showing up in my flex apps. I never had this problem when developing flash apps so I got the idea to use a flash app as a shell with the sole purpose of showing a preloader while the flex app loads. I am not sure how to do this though, and I also need to ba able to pass the FlashVars from the shell flash app to the flex app. Does anyone know how this can be done?
Flex supports preloaders through the Application class, so you shoudn't have to resort to a hack. It's described in detail in the Flex 3 Developer Guide, Ch. 14 ("Application Container"). I never programmed a pure Flash preloader, but I figure it's different from what Flex does. Maybe the two of them conflict?
I am going to begin by agreeing with Simon, but there are ways to create the hack you're talking about.
Unfortunately, the parent layer to the Application in Flex overrides numChildren so that it always returns -1. This means that you are neither able to access the parent swf from flex, nor are you able to directly access the descendant SWF from Flash (amusingly, if a Flex App loads a Flash swf, this is not a problem).
The way around this:
Create a custom event. Set bubbling to true and cancelable to false.
Add an event listener for that event on the Flash layer.
Dispatch that from the Application layer of the Flex App.
The function which handles that event can now use the target property of the event to access the Application layer of the Flex App.