q1ozhv6WqeyPcHMo+ZUlrA88x90LXf+GmMI7aqOmW8n9sIaaA==   
   sI4I2G7Z5qC6/nQl94gWsvtVYEeIaaA32pK85jxl0Sco2RsvyGKAGS4NriuXDhWM   
   oy87fLWinqKtaigiYh/499KUNRnIeQIEe0ipBmBlmvvET19H"; mail-complain   
   s-to="abuse@albasani.net"   
   .os2.ecomstation:834   
   From: "Ruediger Ihle"    
      
   On Wed, 9 Feb 2011 03:49:20 UTC, Steve Wendt  wrote:   
      
   > The claims about acceleration I have seen were primarily for DIVE   
      
   Nope. The acceleration is gained from the use of a shadow buffer,   
   that avoids these awfully slow readbacks from VRAM. That is the   
   main reason, why it is faster than SNAP's VBE mode. If you turn    
   off the shadow buffer, the speed drops noticable. DIVE is not    
   accelerated at all by Panorama.   
      
      
   > Panorama doesn't handle DIVE correctly (which is why there are    
   > workarounds in Mozilla, QT4, SDL, etc.).   
      
   Yep. The problem is, that DIVE bypasses the mentioned shadow buffer.   
   This causes problems, when the same window area is accessed by both   
   DIVE and GPI. I.e. you blit an image via DIVE (single shot) and then    
   drag it around or scroll it using PM functions. I don't know, what   
   Mozilla and SDL do, but for QT4 I can say, that it disables DIVE   
   blitting completely when the presence of Panorama is detected.   
      
      
      
   --    
   Ruediger "Rudi" Ihle [S&T Systemtechnik GmbH, Germany]   
   http://www.s-t.de   
      
      
   --- Internet Rex 2.31   
    * Origin: S&T (1:261/20.999)