If I had a nickel for every time an algorithm was found to have an exploitable weakness due to unforeseen alignments of input, I’d certainly have some nickels.

If I had a nickel for every time an algorithm was found to have an exploitable weakness due to unforeseen alignments of input, I’d certainly have some nickels. We know what the common screwups in crypto are, and we could know what common screwups in network graphs are. I’m just wondering if anyone actually does know of any of those.