PersistenceCapable

My current project is using JPA, heavily. JPA is a spec/standard/interface and I'm using the Hibernate implementation for it, but I import the entities from a different maven module. The entities module has a dependency on JPA, but not on Hibernate. So when I run my unit tests, I suddenly get a: java.lang.NoClassDefFoundError: org/apache/openjpa/enhance/PersistenceCapable Openjpa? … Continue reading java.lang.NoClassDefFoundError: org/apache/openjpa/enhance/PersistenceCapable →

An easier to use Java REST client

I recently was duped (by myself) at a discussion with a colleague: in our continuous effort to disentangle release schedules of two tightly entangled applications, I exposed some core features as a JSON REST service and told him to "just use the Spring invoker" for JSON. It turns out, there is one for JaxRS, Burlap, … Continue reading An easier to use Java REST client →

Integration testing Spring controllers

Time-constrained projects sometimes mandate crude realism as to the type and extent of test coverage in a software project. You'll find arguments for and against testing in the entire spectrum of opinions from "tests are luxury we can't afford" to "regressions are luxury we can't afford". Constrained or not, there is rarely a good excuse … Continue reading Integration testing Spring controllers →

On JPA identifiers and business keys (or why business keys are a bad idea)

JPA, and for that matter Hibernate, depends on identifying entity instances (I'll refer to them as entities for short) by some key. That key can either be a single property (like a numeric ID column in the corresponding table) or a business key (also referred to as natural key). Business keys are entity properties which … Continue reading On JPA identifiers and business keys (or why business keys are a bad idea) →

On coarse- vs. fine-grained synchronization

Performance related topics seem to constantly attract attention and everyone seems to have strong opinions on the matter. Writing code that can be proved [1] to perform according to a certain pattern is probably not the worst place to start with, although the web is full of counterintuitive [2] examples [3] where hardware details weigh … Continue reading On coarse- vs. fine-grained synchronization →