Henry Baker wrote:---
Re keys:
I read recently where someone (at MIT??) had been successful in duplicating a real key from a surreptitious video of the original (traditional mechanical) key in someone's hand. I guess it was relatively easy to figure out the lock manufacturer & then the vertical grooves & then the teeth. Apparently, the number of distinct lengths of teeth isn't very large.
There's also the Matt Blaze paper "Cryptology and Physical Security: Rights Amplification in Master-Keyed Mechanical Locks", available (among other places) at http://www.crypto.com/papers/mk.pdf In an aside in the paper he notes that the number of lengths for teeth on master keys is small enough, and that restrictions on key lengths for adjacent keys further reduce the space, that it is sometimes possible to reproduce a master key from a glance. (That's not the main point of the paper, which is how to efficiently construct a master key from only one copy of a regular key and the lock that it opens (with no operations on the lock other than trying candidate keys)).