Mailing-List: contact flex-dev-help@incubator.apache.org; run by ezmlm
Precedence: bulk
Reply-To: flex-dev@incubator.apache.org
Received-SPF: neutral (nike.apache.org: local policy)
MIME-Version: 1.0
In-Reply-To: <50D39275.7030009@gmail.com>
References: 
 <CAD8jDGL2Fu_sdkHq1Juism5P1jCTeY92SA2Np2fCN=-TcgTG1Q@mail.gmail.com>
	<CCF8A5AB.4A53F%aharui@adobe.com>
	<CAD8jDGLGpXzjLUht=vmjLWR3Sv+jqsdhpYjY51ec2Su5iR=DYA@mail.gmail.com>
	<50D387F0.9000606@gmail.com>
	<CAD8jDGJBHr5qg70KQp0JTZk_9bTyaiA7EsQusyPu8-WP-64phA@mail.gmail.com>
	<50D38E9E.40907@gmail.com>
	<CAD8jDG+7UyPwJnPYBi_7CNafqezZ0P-96MmxgUHoUOz9P5VtUw@mail.gmail.com>
	<50D39275.7030009@gmail.com>
Date: Fri, 21 Dec 2012 00:00:20 +0100
Message-ID: 
 <CAD8jDG+w1VHwY3RnUOrMT3ZrGyV9KVEPc9M3N4M_Br=Wr6uY5Q@mail.gmail.com>
Subject: Re: [FalconJx] New JavaScript runtime format prototype
From: Frank Wienberg <frank@jangaroo.net>
To: flex-dev@incubator.apache.org
Content-Type: multipart/alternative; boundary=20cf3074b9fafa1e9504d150b173

--20cf3074b9fafa1e9504d150b173
Content-Type: text/plain; charset=ISO-8859-1

On Thu, Dec 20, 2012 at 11:34 PM, Daniel Wasilewski <devudesign@gmail.com>wrote:

> I don't know why but it looks like it iss adding extra stars:
>
> http://fwienber.github.**com/**as-js-runtime-prototype/<http:**
> //fwienber.github.com/as-js-**runtime-prototype/<http://fwienber.github.com/as-js-runtime-prototype/>
>
> I obviously realised that, and was able to see your sources,


Sorry again for the garbled link. Using gmail...


> but it didn't change my mind about impact on performance using RequiredJS.
> It is putting you down to JQuery level pretty much.
>

See above, I disagree. JQuery introduces considerable runtime overhead,
even while the application is running.
Again, I propose to use RequireJS for class linking only, so for a normal
application, all define()/require() callbacks are done during application
start-up and then never again.
We would have to create a performance comparison that takes into account
that a class is only loaded once (involving a require() callback), but
instantiated many times without that overhead, and also that an application
does usually not consist of ten-thousands of classes (while
ten-thousands of objects are not unlikely at all). The complex enterprise
RIA we develop runs with a bit over 3000 classes.
Even if using AMD would increase start-up time by some linear factor, do
you really think start-up time / performance is so crucial that you would
sacrifice a clean dependency management that helps avoiding global
namespace pollution and allows even for circular static dependencies?

-Frank-


>
> Dan
>
>
>
>
> On 12/20/2012 10:27 PM, Frank Wienberg wrote:
>
>> Oops? Can you give me a pointer which link does not work? I can't
>> reproduce.
>> Maybe the new subdomain created by GitHub pages (fwienber.github.com) is
>> not "spread around the world" yet?
>>
>> On Thu, Dec 20, 2012 at 11:18 PM, Daniel Wasilewski <devudesign@gmail.com
>> >wrote:
>>
>>  Btw, yes I have checked your links, but stuff on github is 404.
>>>
>>
>

--20cf3074b9fafa1e9504d150b173--