CompuServe Messages

Run up an 8 bitplane workbench and let it scroll. It is, I repeat, a bleeding dog. That is the simplest test to make. When you run up enough bitplanes in non-interlaced modes that you have used the entire memory bandwidth in fast ram things turn very slow very fast. They always have. Reduce things even a little so that you get some of the CHIP ram access back and things pick up speed quite nicely. Do your calculations for vectors outside of CHIP ram and simply draw lines and I expect things to be quite fast. I still expect floodfills that alter several bitplanes of an 8 bitplane hi-res-laced screen to be slower than on a 6 bitplane lo-res screen for example, much slower. {@_@}