Minh wrote:...then you've got the WMF exploit. If I see another "buffer overflow that would allow remote code execution", it'd be too soon.
Actually, I believe that the WMF hole wasn't the result of a buffer overflow, it was "broken by design" in that the security hole was actually a documented WMF feature.
You are correct. But, on the other hand, the fact that it was a documented feature makes the exploit MORE creepy, not less.