We're thinking about it but no immediate plans.  We don't need a common 
location for caching, so I think the need is reduced a little bit.  I also 
wonder what those JS libraries do, in theory if they're loaded from different 
domains they're not allowed to interact with the HTML page due to security, 
right?

We are thinking about ways to have the SWZs get out there even more 
aggressively, but it's likely to be in the Flex 4 timeframe.

Matt

On 9/8/08 2:41 PM, "Doug McCune" <[EMAIL PROTECTED]> wrote:




A related question for Matt or someone from Adobe... didn't Google
announce sometime back that they are hosting common JavaScript
frameworks? The idea being that people could link to those to ensure
that download speed was super fast and that they could be cached more
easily?

Any thoughts on Adobe hosting a version of the framework RSL on a
super duper balls to the walls fast network, so we never have to worry
about the 500+k loading being a bottleneck on our end?

Doug

On Mon, Sep 8, 2008 at 2:36 PM, greg h <[EMAIL PROTECTED] 
<mailto:flexsavvy%40gmail.com> > wrote:
> Hi Doug,
>
> Thanks!
>
> My apology for not including in my initial post the analysis that you have
> so graciously provided.
>
> In our case, we have done the analysis that you indicated, and yes there are
> cases where visitors downloads will be increased by our use of the Framework
> RSLs.
> -- First Time Visitor.  First time visitor for whom our app is the first app
> using that version of the Framework RSL.
> -- Flash Player Earlier Than 9.0.115.  All visitors using earlier versions
> of Flash Player than 9.0.115 will download the swf version of the RSL in all
> cases.
>
> The above being understood, we have decided to proceed with using the
> Framework RSL because repeat visitors, and hopefully even many first time
> visitors will enjoy smaller and therefore faster downloads.
>
> Regarding monkey patching we have not done that in this app.  We certainly
> will be mindful of this if we should we, in the future, do any monkey
> patching (aka "underriding" core framework classes).
>
> Again, thank you for sharing.
>
> g
>

    

Reply via email to