CompuServe Messages

Haven't benchmarked it yet. I suspect that's best left up to someone a bit less interested in the results than I am. C-: I think the best benchmark for a cache would be with some practical application, like compiling a whole bunch of modules that will be re-reading a lot of files like the #includes, but also will keep reading and writing new different files that would flush out retained cache buffers. Come to think of it, though, I've been running with a real large cache instead of using a RAMDISK so buffers practically never get flushed. But for 512K and 1Meg systems, it's more important to retain the right buffers than it is to have all out blazing speed. …cheath