There is no such thing as an int32 in the C++ standard. In this project, as you've seen, it is a typedef, or alias, for unsigned int. It just happens to be 32 bits on the platforms the code is intended to compile on. A compiler could generate code using 64 bit integers and still be 100% standard compliant.
The packets sent between server and client are the result of packet capturing and reverse engineering. It is possible that some of the types used are not exactly the same as what the client expects, but the sizes of the packets are almost certainly correct which means changing values to a smaller type would result in an incorrectly sized packet and also offset anything that came after that value unless you add the appropriate amount of padding to match the size. If you believe a packet is being sent incorrectly the right thing to do is capture traces of those specific packets and check them against the existing code. Perhaps the packet you're looking at does expect 0xffff and not 0xffffffff in that position, but the only way to know for sure is to verify it against packets from live.
|