The
efnx code blog.

Interfaces in AS3 are super useful for requiring classes to support a set of public methods, enabling other classes to communicate through the interface thereby abstracting the class itself. AS3 doesn’t natively support abstract classes so using Interfaces is a good replacement in some cases, but if you’ve never used an Interface before you probably wouldn’t know how to implement it. The docs help a little, showing you how to instantiate a class using an interface with the “implements” keyword, but doesn’t go much further:

But what if you’d like to implement a class while extending another? Well, it’s simple once you know the syntax…

This is a very quick and simple tip that took me a while to find online due to the fact Adobe doesn’t mention it. Such a simple thing should be in the help docs, but I guess adobe thinks that anyone trying to use Interfaces would already know the syntax! Well now you do.

I just like posting vidyos when I’m too lazy to code or write about code.

[edit august 21st 2008] Hey guys this code is available but it is messy and the project has been scrapped. I will pick it up again, probably with OpenGL and C++[/edit]

In an effort to (maybe) find a new job, depending on what happens with my current one, I’m releasing a preview of my GUI classes. This class is a window class. It contains your display objects in a nice, draggable, scrollable and real-time skinnable window. You can make your own graphics for the window and configure button placement, scrollbar offsets, label, offsets and all other text goodies from an external config.xml script. There’s nothing to keep in your library and it’s not that big. It would be nice if anyone who uses it can tell me their memory stats, as I’m currently looking to make sure my code is clean and there are no leaks. Click the link below for a little demo.

Here’s the demo!

If you’re interested in using the class you can dive in headfirst at your own risk. It’s kinda long (about 1200 lines). I’ll be working on completely revising it from the ground up as it really is a big mess right now.

Source: GUI – mess

While working on my windowing system I ran into a problem. Every time I tried to set the width of my window class instance I’d get an error from my bitmap resource handler telling me the resources hadn’t been loaded. This makes it impossible to set widths for a window before it’s loaded, and since the entire system is supposed to be deployable online, loading time is not anything I can count on. I wrote this small class to take care of executing functions at a given time (like after loading is finished). The class is called Queue and here is some usage:

That code does this -> First we define two different functions, blah and blah2. Then we push the functions to be executed into the queue. Here’s the interesting part. Since the Dictionary Object inside the class used to list functions is not weakly referenced Duplicate entries cannot be made by the same function regardless of it’s parameters. In the the example above I push blah into the queue three times, but the only one that fires and traces is the third because the queue’s key is the function itself [or the "closure object (ie. the "behind the scenes" object that facilitates method closure by maintaining a reference back to the method and its scope)" as described by Grant Skinner], so the function replaces itself in the queue. This is nice for doing things like setting widths and resizing, although not so great for things like sending different strings to some handler. As of now calling start() applies the functions asynchronously and nullifies anything left in the queue. You’ll see the output traces this:

First it traces the Queue Object which contains references to the two functions and their parameters (not in the order we gave them) then the output. Bam! Done. Oops! Not quite… …At any point you can flush the queue without executing the functions by calling abort(). There, done.

This class can be easily extended or modified to include directives to execute functions in a given order or multiple entries of one function, so get crackin!

Source
Queue Class Source Code

I’m now using a new class for this task which can be found here. That said, feel free to read on about this outdated class.

I’ve completely rewritten my multiple bitmap loader into a new machine. It now uses weak referenced dictionary objects to store bitmapData after it has been extracted or loaded from the web. It is meant to hold all instances of bitmapData for a class, a set of classes or a whole game even. The class, bitmapDataRsrc, takes either a bitmap, bitmapData or string path to an image or an array consisting of either bitmaps, bitmapData, string paths, or some combination thereof, along with a string identifier and loads the bitmapData into a dictionary object with it’s identifier as a key. It then pushes it’s load status [true or false] into another dictionary with the object itself as the key, so every bitmapData resource and it’s load status is accessible through the identifier. Once every resource is loaded bitmapDataRsrc dispatches an Event.COMPLETE event. For instance, let’s say I wanted to load and store all the bitmapData needed to build a GUI window:

After the resources are done loading you can access the bitmapData using the get() function with the string identifier [2nd parameter] passed with the load() function. Below I’ll demonstrate accessing the last bitmapData, XbuttonOver.png.

If you get lazy and decide not to give bitmapDataRsrc string identifiers to identify your resource, it will make one up for you. If you decide to pass an array of bitmaps, bitmapData and string paths to images as the first parameter to the load() function, and pass a string identifier as the second parameter your resources will be indexed as “string identifier + number” so that all your resources are accessible. When passing an array of resources you can identify each one yourself by first indexing the resource and string identifier in another array and then pushing that array into the parameter array. In this case a second parameter to the load function is not needed. Check it out:

I find it a lot easier to load resources one by one and name them, it’s a lot better for me, I like to get to know my bitmapData, but it’s good to have options.

The isLoaded property can be checked to find out if all resources have loaded. At any point the load status of a single resource can be found by calling the hasLoaded(“string identifier”):Boolean function. If no identifier is given bitmapDataRsrc returns it’s isLoaded property. To destroy a given resource call unload(“string identifier”) and to destroy all resources in the manager call unloadAll():

Source
bitmapDataRsrc Source

Scott Peterson has created a tool for porting C/C++ code to AS3. What the shit. That rocks. This demo was shown at the 2007 Adobe MAX conference, in the sneak peeks section, so keep in mind it may never see the light of day. But how cool is this? Adobe better jump on this.

This class has been updated and a new post has been made, click here for that post.

Also known as “Developing a Pure AS3 GUI – Part 2 (buttons)”

Buttons are extremely important in a GUI. Buttons don’t have to be fancy to be useful, they just have to execute some function when clicked. Although, seeing as this is a button class for AS3 and AS3 is the language of Flash and Flash is by it’s namesake, “flashy,” this class is somewhat “fancy” beyond the utility of a simple button. Don’t be discouraged if you’re looking for an easy implementation, by no means is this class complicated, it just goes further than binding a Sprite to a function using a CLICK event listener…

My button class requires three images to function: an “up” image [the default resting image], an “over” image [the image shown on mouse over] and a “down” image [the image shown when clicked upon]. After creating the three images and placing them in some directory next to your .fla you can instantiate the button by giving it an array consisting of three image paths that lead to your images. Additional parameters are a function to be performed on click and whether or not the button is a toggle button.

Usage

And that’s it. If you’d like the button to run in toggle mode just pass true as the third parameter.

Source
button Class Source Code

To aid in my projects I have written a simple bitmapData loading class entitled aquireResources. This class takes an array of image paths as strings and proceeds to load them one by one. A “complete” event is dispatched when loading is completed and the bitmapData of each loading instance are saved in an array in the order they were given. Calling the aquireResources.array() method returns the array of bitmapData. If you’ve already downloaded Thibault Imbert (bytearray.org)’s GIFPlayer and GIFDecoder classes you can uncomment the included properties and methods in my class and enable the aquireResources class to load multiple instances of animated GIFs into AS3 as arrays of bitmapData. In this case aquireResources.array() returns a two-dimensional array containing one array of bitmapData per animated GIF passed to it.

Usage

For loading animated GIFs

Source
aquireResources Source

Also known as “Developing a Pure AS3 GUI – Part 1.5 (tile)
Jump to Source
After writing last night’s class, scalingBars, I decided to make it a little easier on myself to tile bitmaps across the screen. I wrote a simple class that when given either a BitmapData object or a path to an image file (as a String), class will tile the image x pixels in width and y pixels in height. This is great for backgrounds or window bars or box sides, etc. Here’s the usage:

The previous code will grab an image from “pathToImage” [either relative or absolute, but remember security!] and then tile it 100 pixels wide by 150 pixels tall. Simple. Now let’s say you wanted to resize it:

That code will re-tile the image to 20 pixels by 40 pixels. Next on my list of upgrades is to allow reloading of different images. Shouldn’t be too hard.
Example

Source
tile Class Source Code

This class has been deprecated and phased out as of 12.7.2007

Writing your own GUI can be a daunting task, most people might find it easier to learn MXML and use Flex Builder or Flash Components than to waste their time writing GUI code and Photoshopping images. I on the other hand hate proprietary frameworks [like MXML] and really like the clean approach of creating classes of GUI elements that build on each other to make an interface. This post is the first in a series of posts designed to step through MY method of GUI design.

Example
Source

I start the process at windows. Windows hold all the content of a graphical operating system, the mother of all interfaces, so why not start there? Well, windows are super complicated, so first we’ll start by writing the components of a window. We’ll need a window bar that holds the close, minimize and optimize buttons and scales to a given size. This means we’ll need bars [vertical and horizontal] that tile. Don’t let the name Scaling Bars trick you, these bars aren’t really scaling, they’re tiling the bitmaps given to them in the constructor. Here’s the beginning code where we set up all our variables.

Okay. We’ve imported some classes that I’ve written myself that I haven’t talked about before, don’t worry, I’ve included them in the source and will post about them later. The vars are all pretty self-explanatory. The loadedSprites are a class of Sprite I’ve developed to make importing graphics a little easier on me. Given a path to an image it simply retrieves the image and sticks it inside a Sprite [itself]. On to the constructor:

When instantiating the scalingBar class, create an array consisting of paths to the three images to be used for the bar, the left [or top], the middle, and the right [or bottom]. If left to the defaults your bar will tile and size horizontally. You can pass “vertical” as the second parameter to set the bar up to tile horizontally. Next we have the functions that take care of placement and size checks after the images are loaded. loadedSprite operates whatever function was passed to it as the second parameter in its instantiation, so no external event listeners are needed. We passed loadCycleComplete to the loadedSprites so that function will be called each time one is done loading:

Almost done! The last part is the resize function, which takes an integer size and scales the entire bar either horizontally or vertically, depending on what you have specified. This is the real meat of the class. The resize function moves the bottom [or left] image to the border of the given size, then tiles the middle image over however many times needed until it reaches the bottom [or left] image:

Example
And there we go! Here is an example of the implementation:

Source
scalingBar source