Click on any phrase to play the video from that point.
[Adobe® Flash® Platform]
[♫ upbeat music ♫]
We created SonicSwap because we saw
an inherent advantage in Web-connected music--
that it would be easier to discover new music if I could see what's in my
friends' music collections, see what my friends are listening to,
see their playlists.
SonicSwap, in seconds, allows you to upload your entire iTunes metadata--
all of your playlists and everything that you listen to in iTunes--
up into the cloud.
Not the actual tracks themselves, just the information
about what's in each of those playlists, and then allows you to access those playlists
on any platform by finding sources of music on the Web.
When we started to build SonicSwap two years ago, we used
conventional Web technologies that were available at the time--
standard HTML, MySQL, etc.
We were unable to process HTTP requests in any reasonable amount of time.
Servers were crashing.
The other problem is that the database got humongous.
For a very few number of beta customers, our databases got so large
that even MySQL's best consultants told us something had to change.
SonicSwap's site was, frankly, dead in the water.
Those scaling problems were going to be the death of the company
[unless] we could solve them and solve them quick.
We made the decision to do the cutover to the new architecture
using the Flash® platform in late June.
Peter Armstrong's presentation on flexible rails
gave me inspiration to move much of our Web application software
out of the server's side and into the client's side.
Every customer that comes to SonicSwap and TuneVision comes with a computer.
They come with an Internet connection. They come with an IP address.
And so we could leverage the resources that they bring to the model to make this
infinitely horizontally scalable.
We can then do mash-ups with other third parties.
For example, we go out to YouTube and use their services, BlastFM services,
Amazon, Google's services, outside of YouTube that allow us to enhance
the metadata right there in the Flex® client.
As a developer, I like Flex® for many reasons.
One, it's open-source and you have the ability to customize the components
and customize the way things look and feel.
Using the Flash® platform, we can really focus not on the back-end code,
but instead what's really important, which is the user experience.
As we started to develop more and more Flex® applications,
and we understood exactly how to leverage the power of Flash® and Flex®
at the user level, we discovered that we could automate playlist generation
for the end user right on their desktop.
TuneVision was an extension of SonicSwap.
It allowed us to create a user experience not unlike Pandora,
where you enter in the name of an artist and we build the playlist on the fly.
And we could leverage the data stored on SonicSwap of your favorite artists
to also go off and build an ideal artist video channel
that's a very personalized music experience on the Web.
[♫ music ♫]
Well, working on new applications that allow us to mix AIR® technology
with Flash®, where we have a Flash® playlist, if you were running on a Web application,
and an AIR® application running on the desktop.
When the user loads a playlist from one of their friends, if they have the track
they can play it off their local drive.
And if they don't have the track, we can mash it up with other music sources
on the Web to give them access to that music.
None of that would be possible with any other architecture other than
a mixture of AIR® and Flash®. Nobody else could do that.
A lot of the growth opportunities today are really happening in the mobile platforms.
By using Flex® and Flash®, it's actually enabled us to move onto those platforms
very, very quickly.
Flash® 10.1 on the mobile platform has opened up the entire mobile market to us.
The same staff that is now supporting what we're doing on the Web,
is also allowing us to support these additional platforms
with very little development cost and an enormous amount of
reuse of the code and leveraging what we already have.
A lot of bloggers out there were saying that Flex® was not a viable solution
because Flex® was heavy, and it produced code that was too big for a mobile version.
We put it on the Android Nexus One. To my surprise, it worked extremely well.
It was faster, more snappier than the same program on a browser on a normal PC.
The fact that these Flash® plug-ins are in some mobile phones--it was a lucky break.
We call it the lottery for us.
Flash® has been critical to our success.
We would be dead today if we had not modified our architecture
and included Flash® and Flex® at the desktop level to offload a lot of the
functionality onto our end-users' computers.
Today users commonly have a much, much better user experience
because we can leverage the power of their computer, their Internet connection,
their bandwidth to deliver a richer music experience.
None of that would have been possible if it weren't for the Flash® platform.
[♫ upbeat techno music ♫]
[Adobe® Flash® Platform]
