Proposition: change vector_size_t from signed to unsigned

[czentgr]

The vector_type_t is defined as int32_t

velox/velox/vector/TypeAliases.h

Line 29 in 4cffa4b

using vector_size_t = int32_t;

In the context of vectors, the std::vector class refers to the size type as unsigned integer type (usually std::size_t).
vector_size_t type variables are also used as an index into the vector. And it doesn't quite make sense to allow this to be negative.

I'm sure there was a reason at some point to use signed it. The comment says this type was created to unify the usage so it wouldn't be uint32_t and int32_t all over the place. But why was it not uint32_t in the first place?