Today i want to implement an easy contact form in one of my older projects but there was a strange behaviour.

I have created 2 new instances of LoadVars class and set a release event on to a button for calling a sendAndLoad event.

[as]
public function init():Void {
mailInfo = new LoadVars();
mailStatus = new LoadVars();
mailStatus._container = this;
mailStatus._container.sucField = _container.sucField;
mailStatus.onLoad = onMailLoad;
}
public function release():Void {
mailInfo.fVorname = _container.forename.text;
mailInfo.fNachname = _container.surename.text;
mailInfo.fEMail = _container.mailadress.text;
mailInfo.fFeedback = _container.feedback.text;
mailInfo.sendAndLoad(“kontakt.php”, mailStatus);
}
[/as]

This is the callback function

[as]
public function onMailLoad(success):Void {
if(success) {
_container.sucField._visible = true;
_container.sucField.gotoAndStop(1);
} else {
_container.sucField._visible = true;
_container.sucField.gotoAndStop(2);
}
}
[/as]

The callback is defined in the same scope where loadVars are defined.

The problem:

After compiling fla-file and testing contact form, i recieved email and get answer from server(i tested with packet-sniffer Etherreal), so i thougt, what the hell is wrong.

The next try i set an additional debug textfield in my fla and trace the reference to the class where the callback is defined and to the output field which is set visible in callback function.

Both where undefined and so thought about how it was in Action Script 1.0 . There you have to create an anonymous function if you wanna get a callback, or setting an other function to overwrite the LoadVars onLoad.

Example:
[as]
var l = new LoadVars();
var l2 = new LoadVars();

l2.onLoad = myFunction;

l.sendAndLoad(“test.php”, l2);

function myFunction() {
trace(“callback”);
}
[/as]

With this construct the scope of the function was to the LoadVars object l2.

In ActionScript 2.0 there is the same behaviour, the function onMailLoad will be called, but with another scope. It is not longer member of the class where the method is defined, but the method is member of loadVars instance.

very strange

An Obfuscator is a tool which allows you to save your bytecode from the decompiling gang.

After obfuscating a precompiled file under normal circumstances it is not possible to decompile.Why?

Not all Obfuscator tools works similiar or produce exactly the same output, but the concept is nearly the same. One idea is to rename variable names to constructs like this; “ooooooooooooooooooox”, no one will be able to read source code like this. But a good decompiler maybe could reproduce the code. So what can a good obfuscator do?. A solution is to rename source via a mathematical formular and a key-string.

The developers of the Motion-Twin Action Script Compiler have a page on their website where they serve a tool for irreversible obfuscating. The only possibility is maybe a brute-force attack, but only for short variables.

The tool is called OBFU (for OBFUscator).

Obfuscation process is irreversible. Every identifier string found in the SWF is hached with a irreversible proven mathematical function. For short variable names however, brute force is still a possible attack. In order to prevent this Obfu accepts an obfuscation key as parameter (using the -key parameter). Two SWF will be obfuscated in the same way if and only if they have the same obfuscation key. In order to ensure maximum security, one should use -rndkey to generate a random key everytime an obfuscation is done. If however several SWF needs to communicate between each other, there is several possibilities :

* use the same keys (with -key) for the two SWF. This way they will be obfuscated the same way and thus a function f declared in the first SWF and called in the second will be correctly executed.
* protect the variables that need to be shared so they don´t get obfuscated (see below).

Sounds good, but there is only one problem, professional licence costs are about 1500 $.

Maybe only for agencys, but if you have seen the results in the sample swf, it is maybe rentable.

If you are a satisfied SEPY user you maybe wanna check the synatx rules inside the editor and not to flush time by time.

Now it is possible with the open souce Action Script 2 compiler MTASC.

In newer SEPY versions there it is an easy thing to implement the MTASC compiler.

In SEPY IDE there is a button, similar to the check synatx button in Flash IDE. If you press and there are some errors, SEPY will go to the line where the error occured.