v2.2

Relax memory ordering: use std::memory_order_relaxed everywhere in current implementation

v2.1

• Implemented custom allocators which are different for different threads (but they are still using the same global allocator inside)
• Code cleaning
• Thread safe hash objects

v2.0

First lock-free implementation of key-value web server.