I've noticed that the convention for JMX MBeans appears to deviate from the standard Java Bean property model in that the names for attributes will traditionally start with a capital letter, i.e. PascalCase.
To explain this a bit more clearly, I'll take an example from the JDK (chopped down a bit for clarity) :
public interface MemoryMXBean {
public int getObjectPendingFinalizationCount();
public MemoryUsage getHeapMemoryUsage();
public MemoryUsage getNonHeapMemoryUsage();
public boolean isVerbose();
public void setVerbose(boolean value);
}
And the Memory MXBean exposed has attributes V erbose, H eapMemoryUsage etc.. This is also the case for every MBean exposed, and you receive the same behaviour when defining your own MBeans.
Does anyone know the origins of this convention / implementation detail? I glanced through the JMX spec and only found a passing referencing to case sensitivity of attribute names, with no clear convention mentioned.