ORM is a neat idea, typically used together with rich domain model. JPA, currently in version 2.1, is a Java specification that is now a common ground for most Java ORMs around. But ORM is also a love-hate topic, with stickers like "Vietnam of computer science", often called an antipattern.
In this book we'll talk about many interesting JPA aspects and put them into the context of other persistence options. I'll also try to explore use of JPA 2.1 in a way much closer to SQL, but still benefiting from many features that make it so convenient when compared to plain old JDBC. We will discuss non-JPA feature called entity (or ad-hoc) joins - supported by both EclipseLink and Hibernate. We'll also cover some typical problems (lazy loading and related N+1 select) and some less typical ones (like modularity).
Querydsl will be with us along the way, because compared to Criteria API it allows us to write much cleaner queries in a fluent API fashion.
I have been reading the not yet finished book. If you already know something about JPA, you may consider this as a guidepost representing some ideas. It is kind of a distilled blog but not enough to be a decent book.