Working on my current project I’ve faced a very interesting situation. For each object of a given type, I had to create a monotonically growing identifier with few caveats: 1) the solution should work in multithreaded environment 2) the number of objects is fairly large, … Continue reading Managed object internals, Part 2. Object header layout and the cost of locking