How the Flash Search Player works

Jim Corbett, Flash Player Engineer at Adobe, did a talk at MAX San Francisco with some good information on how the Flash Search Player works.

http://tv.adobe.com/#vi+f15384v1000
45min video

What I got out of it:

• Adobe has built a headless flash player that runs through a
  swf file really fast.
• Google has build a virtual user that communicates with the
  headless flash player, the user deals with the logic to make sure it
  clicks all the buttons and decides what to index.
• All text fields visible on the display list can be accessed
  even dynamic ones
• Images of text are not found – best way to hide
  text for Easter eggs
• If you load data via external file i.e. xml and populate
  text boxes, it can be accessed via the headless flash player (Google
  are not yet passing this file back to the headless player, but they do
  index the external file)
• No sockets, no video
• A snap shot of the state can be taken (this is at the
  Google virtual users request)
• No external interface
• No swfAddress deep linking
• Flashvars are ok as the virtual user passes the player the
  complete url
• No accessibility text yet

This entry was posted on Monday, December 15th, 2008 at 11:26 am and is filed under Flash, Flex. You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.