i've tried to experiment porting vblade-20 to win32 native using multithreading, winpcap (sending/receiving packets) and createfile/setfilepointer/readfile (replacing direct ata calls), i've just barely made it work, but i'm a bit disappointed about the speed, there has been a vblade ported to cygwin that i am also using and that is much more faster then my ported vblade...i'm wondering why...

does direct ata calls are signicantly faster than emulating them through readfile?

does linux native putpkt faster than sending through winpcap? (been using winpcap que function to que sending packets)

i've been porting this project using orwell dev c++ and using multithreading so that instead of having 1 process/vblade server, i have the option of sticking 10 threads/vblade server...

also i haven't begin to optimize (yet)...