When a patch is released for buggy software, bad actors may be able to analyze the patch and create an attack on unpatched machines. A wrapper could block attacking inputs, but it, too, gives attackers critical information. An opaque wrapper hides such information if it can be implemented. Hashing and Bloom filters can only implement opaque wrappers in very special circumstances. We describe Zero-Knowledge Proofs, Fully Homomorphic Encryption, and Oblivious Transfer Protocols to show that they cannot help, either. Although we see no way to implement general opaque wrappers, we hope this idea spurs research.