"Dmitry Timoshkov" dmitry@baikal.ru writes:
I'm not opposed to your fix, not at all. I just would like to understand better what is the real problem with our code. If you, or somebody else, could confirm (by a test case) that exactly (MFT_BITMAP | MFT_SEPARATOR) combination should be rejected then I'll vote for your patch.
At least on NT4 it's not rejected. In fact NT seems to accept any combination of flags. I have removed the check for now.