Summary: a short intro to reducing DB lock contention with "SELECT ... FOR UPDATE OF x FOR SHARE OF y", and the upcoming Postgres 9.5's "SELECT ... SKIP LOCKED".

It's a decent heads up about these features, but there are some editing errors and an odd emphasis on multiple CPUs, while DB row locks are the limiting factor in the examples. I'm having trouble interpreting the final paragraph. ("Keep in mind: If you only got a single CPU, concurrency will be a problem. Therefore is essential to do things in a way that many CPUs can do their share of work at the very same time.")