The previous patch was working for browser-based attacks, however scripted software was still able to cause an overflow.
The socket.close() function did not seem to act quickly enough in this case. I was wrong earlier about not being able to flush the socket buffer & have added that into this patch. In local tests on 2 different machines (using a software-based attack) this patch is tested and working (again).
The previous patch was working for browser-based attacks, however scripted software was still able to cause an overflow.
The socket.close() function did not seem to act quickly enough in this case. I was wrong earlier about not being able to flush the socket buffer & have added that into this patch. In local tests on 2 different machines (using a software-based attack) this patch is tested and working (again).