No idea, but I suppose it does, because I *can* get through with the simple world. It shouldn't even be a multithreading-related issue, since a CS is locked prior to using the crypto lib's encryptor and decryptor.
If it is a problem in the crypto library, we'll see it because it will fail the MITM attack on the protocol that I'm writing into the proxy.
I committed my current WIP 1.3.2 protocol to the svn, so that anyone can have a look at it.
There's a commented-out MilliSleep in cProtocol132::SendChunkData(); uncommenting it sometimes makes the client accept even the good world, but eventually the connection lags too much behind and the client times out.
If it is a problem in the crypto library, we'll see it because it will fail the MITM attack on the protocol that I'm writing into the proxy.
I committed my current WIP 1.3.2 protocol to the svn, so that anyone can have a look at it.
There's a commented-out MilliSleep in cProtocol132::SendChunkData(); uncommenting it sometimes makes the client accept even the good world, but eventually the connection lags too much behind and the client times out.