debris
Accidental performance regression
This PR renaming to_vec to into_vec accidentally introduced a silent performance regression between 0.8 and 0.9. Since ElasticArray implements Deref<Target=[u8]>, all the usages of ElasticArray::to_vec in the code have now silently been replaced with <&[T]>::to_vec, which is consistently O(n) (whereas ElasticArray::to_vec is either O(1) or O(n) depending on the size of the array). I discovered this when I migrated to 0.9 and the compilation still succeeded. I don't know there's much we can do about this other than documenting it, but it is something that I'd say needs to be documented.
It should be noted that this doesn't actually affect the Parity codebase, since I just ran it through sed and then fixed the places where it really did need to be to_vec. Maybe it would be worth pinging our reverse dependencies.
