This is version 2.12 of my FastWaitBlit patch, which speeds up
multi-tasking,  and  so  isn't  too benchmark-able.  You *May*
experience  lower benchmarks on certain graphics operations in
(for  example)  AIBB,  but  if you run lots of tasks, and your
main  task  isn't  too  graphics-intense,  then you won't wait
around as much.

People  have  told  me  that  my patch isn't as fast as *Art's
FasterBlit  patch, the reason for this, is that my patch works
in   a  different  way  to  *Art's.   His  increases  graphics
performance,  and  mine  increases  multitasking  performance.
Take your pick.

Or, you can run the Non-DMA version of my patch in conjunction
with *Art's FasterBlit, which will use an optimised version of
the  graphics  library WaitBlit() function, without my Dynamic
DMA(tm) idea.

 PERMSSN    UID  GID    PACKED    SIZE  RATIO     CRC       STAMP          NAME
---------- ----------- ------- ------- ------ ---------- ------------ -------------
[generic]                  385     460  83.7% -lh5- e924 Mar 28  1997 FastWaitBlit
[generic]                 3209    7134  45.0% -lh5- 7764 Mar 28  1997 FastWaitBlit.Doc
[generic]                  369     448  82.4% -lh5- c865 Mar 28  1997 FastWaitBlit_NODMA
---------- ----------- ------- ------- ------ ---------- ------------ -------------
 Total         3 files    3963    8042  49.3%            Mar 28  1997