A LazyRc is an Rc but with slightly different tradeoffs. An Rc<T> puts the reference count next to the referenced object in the same heap allocation, while a LazyRc<T> puts them in two separate allocations (only allocating the reference count when cloned the first time).

1. Use an Rc if you have a stack allocated object or if copying the object is cheep.
2. Use a LazyRc if you already have a boxed object, are unlikely to share it, but want to be able to share it without copying it.

For example, if you have an API that returns a potentially large buffer in a Vec<u8> or a Box<[u8]>, you may not want to convert this to an Rc<[u8]> as that will copy the entire buffer. Without this library, you'd have to have to use an Rc<Box<[u8]>> however:

1. That requires additional pointer chasing to dereference.
2. That will always allocate even if the Rc is never shared.

This library will allow you to convert that Box<[u8]> into a LazyRc<[u8]> with no allocations until it's cloned the first time and no copying of the underlying buffer (ever).