View Single Post
Old 08-22-2011, 01:44 PM   #28
kindle3zeng
Enthusiast
kindle3zeng once ate a cherry pie in a record 7 seconds.kindle3zeng once ate a cherry pie in a record 7 seconds.kindle3zeng once ate a cherry pie in a record 7 seconds.kindle3zeng once ate a cherry pie in a record 7 seconds.kindle3zeng once ate a cherry pie in a record 7 seconds.kindle3zeng once ate a cherry pie in a record 7 seconds.kindle3zeng once ate a cherry pie in a record 7 seconds.kindle3zeng once ate a cherry pie in a record 7 seconds.kindle3zeng once ate a cherry pie in a record 7 seconds.kindle3zeng once ate a cherry pie in a record 7 seconds.kindle3zeng once ate a cherry pie in a record 7 seconds.
 
Posts: 35
Karma: 1566
Join Date: Nov 2010
Device: kindle 3wifi
Quote:
Originally Posted by kranu View Post
A 0.5 second delay is pretty good. That's about how quickly you can turn through pages on the Kindle itself. Have you released the program yet?

I've experimented with screenshots myself, but the delay is about 6 seconds, even when connected over USB. I'm very interested in how your difference algorithm is implemented.
The program is pretty rough right now. I might publish it when it's got polished. The key algorithm I used is:
on host,
1) convert pages to kindle framebuffer format, which would be 800*600/2 bytes long,
2) xor two consecutive pages as diff,
3) zip the diff, send over wifi.

On kindle side
1) unzip diff
2) xor the current framebuffer with the diff
3) find out the dirty area, this is where I don't have good algorithm
4) update the screen (only the dirty areas or the whole screen)

Last edited by kindle3zeng; 08-22-2011 at 01:47 PM.
kindle3zeng is offline   Reply With Quote