No subject

even with the optimization of only writing the pixels that arn't white (they 
are all white by default). 

There are other places on the chain of execution where i can intercept the C 
library. The c-lib uses drivers to create the image - the one i was using 
before was an interface to the gd library, the one i am using now writes 
directly to memory. The drivers only have to implement a small set of commands
to satisfy the interface: draw a line, draw multiple lines, and change state 
(color, pen width, etc). Perhaps i could take a bitmap-dc and have the driver 
call the equivelent functions on it directly. Even better would be to be able 
to access the bitmap portion of the bitmap-dc% and simply copy over the bitmap 
(in C) that is created by the c-library without going through scheme.

Any suggestions?

-Alex