The patch adds 'native' floating point support to DGD In addition it adds sin,cos,tan,pow,asin,atan,acos log, log10, exp, atan2 kfuns. You can still use emulated floating point provided with DGD by compiling with -DEMU_FP Your machine must have 8 bytes doubles and 8 bytes long long. (as linux has) The code is alpha and i bet it containts bugs. Person to bug if something goes wrong is sasha@netvision.net.il BE WARNED