>> Have you an idea of what defox IV is doing?

LOL. Yes, but I'm taking Leonid's lead in not provocatively publishing steps to hack! I do not want to create a self-fulfilling prophesy of people roaming around stealing VFP sourcecode. In any case, Leonid's seeding technique makes it quite difficult to predict his varying encryption unless you want to disassemble some very low level routines. Definitely NOT a neophyte task.

Leonid does publish a separate obfuscator you can use before Defoxing, but the heavy encryption is the Defox difference, not alteration to the actual VFP source.

>>I am at this stage NOT using the supplementary "application protection" scheme which is definitely too slow:)

Right, and Leonid makes this point himself and describes how to avoid the slow down. It's because there's heavy line-by-line decryption going on. That's what makes it so secure. In contrast, VFP Compiler eviscerates the VFP project by moving most content to a C++ dll. IME there's no visible performance difference using VFP Compiler and while somebody determined could hook lines of pcode as they execute, Chen's decompostion actually creates illegal VFP verbs that cannot execute, and SCANs and loops are extremely difficult to hack. Again, definitely not a noob task.

But a plain VFP exe inside Molebox or Armadillo or whatever? Anybody reading this could have the project in minutes while standing on your head.