Mailing-List: contact flex-dev-help@incubator.apache.org; run by ezmlm
Precedence: bulk
Reply-To: flex-dev@incubator.apache.org
Received-SPF: neutral (athena.apache.org: 207.97.245.191 is neither permitted
 nor denied by domain of CaptainN@unfocus.com)
Message-ID: <50B54EE4.4010402@unFocus.com>
Date: Tue, 27 Nov 2012 18:38:12 -0500
From: Kevin Newman <CaptainN@unFocus.com>
User-Agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10.8;
 rv:16.0) Gecko/20121026 Thunderbird/16.0.2
MIME-Version: 1.0
To: flex-dev@incubator.apache.org
Subject: Re: FalconJS has landed
References: 
 <CAJs+wW10J6cuNG+vfj33ZNofq+79VZcJP1rVoDF63ejpZ3r-wg@mail.gmail.com>
 <CCD91FE7.49088%aharui@adobe.com>
 <CAJs+wW1O9gh7eTQTR-AsTQyUE4O6JSLpYL+dZqXV+E=fQaio=w@mail.gmail.com>
 <50B4E8C3.5040201@gmail.com>
 <CAJs+wW2iu5UEJFjyO1+1BnV_+nMi93mnen84AJ1pE8PbCtFm-A@mail.gmail.com>
 <50B5052E.6000301@gmail.com>
 <CAJs+wW29ccd1zkXhvXM667b3FY3q_pCYfo4j3rNQJ2GpuTonRg@mail.gmail.com>
 <50B54518.4050204@gmail.com>
In-Reply-To: <50B54518.4050204@gmail.com>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit

On 11/27/12 5:56 PM, Daniel Wasilewski wrote:
>
> However, when comes to FalconJS the task is to use strict proper OOP 
> language like AS3 to translate into JS. In my humble opinion, this is 
> what is all about, to protect us and give us more robust development 
> environment and target the platform we need without writing FOR the 
> platform. The fact that we are using AS3, output doesn't need to 
> leverage all those concepts on the target level at all. Even if the 
> output code would be set of plain objects or procedural style chain of 
> objects, if better for performance and the final result, better for 
> everyone. Good for marketing Apache Flex JS as the robust environment, 
> web RIA development environment.

I agree, and I can't see how you'd emulate certain things in a clean way 
anway - like protected or internal class members. You can kinda do 
privates with closures and literals, but you lose a bunch of stuff, like 
instanceof, it uses more memory, and it's actually slower at runtime 
(even if definition - boot up - is a bit faster).

As for perf tests, maybe throw one of these in there:

function BaseClass( yourVal ) {
     this.someVal = yourVal;
}
BaseClass.prototype = {
     myMethod: function() {
         console.log( this.someVal );
     }
};

function Extended( yourVal ) {
     // super
     BaseClass.apply( this, arguments );

     this.someOtherVal = yourVal + " other";
}
Extended.prototype = Object.create( BaseClass.prototype ); // avoids 
constructing parent
Extended.prototype.myExtendedMethod = function() {
     console.log( this.someOtherVal );
};

var ext = new Extended( "test" );
ext.myMethod(); // test
ext.myExtendedMethod(); // test other

ext instanceof BaseClass; // true
ext instanceof Extended; // true

That's the closest I think you can get to classical inheritance in pure 
JS (you'll need to polyfill Object.create for IE), while maintaining the 
prototype chain.

https://developer.mozilla.org/en-US/docs/JavaScript/Reference/Global_Objects/Object/create

Kevin N.