Unique types are used to guarantee that only a single reference is ever held to a value. Sort of like the value has a single "owner". This restriction, while a

Unique types are used to guarantee that only a single reference is ever held to a value. Sort of like the value has a single "owner". This restriction, while a maybe bit of a pain to program with, gives the compiler permission to do clever things. In particular: 1. The compiler can detect when a value is no longer "owned" (referenced) by anything and free it automatically -- without garbage collection. That's really handy for things like closures where the compiler automatically allocated the memory for you in the first place. 2. If an immutable value is modified then a copy usually needs to be made. But if the immutable value is uniquely-referenced then the compiler can reuse the old bit of memory, thereby saving a copy operation. It can do this because it can prove the old memory can no longer be accessed. 3. I think Rust might also use uniqueness when sending values between its tasks. Since it can prove the value will no longer be referenced by the old task the compiler can avoid copying the values while still preserving isolation. http://en.wikipedia.org/wiki/Uniqueness_type http://en.wikipedia.org/wiki/Uniqueness_type