What about collisions? Even if the hashing algorithm is the same everywhere, two different urls can yield the very same hash, and the birthday effect (that is,

What about collisions? Even if the hashing algorithm is the same everywhere, two different urls can yield the very same hash, and the birthday effect (that is, there's a very large chance that two students share the same birthday in a class of 30) makes that an important problem. URL shorteners need to have some sort of mechanism to deal with hash collisions, but this means that it can only be assumed that two identical hashes point to the same url if we know that the very same database is backing both. Sure, you can increase the length of the hash to a level where the birthday effect isn't very significant but that would mean longer hashes. For hashes only five-six characters long, the chances of collisions are just too high.