Joachim Sauer created DAEMON-309: ------------------------------------ Summary: Documentation for start method in JVM mode is conflicting Key: DAEMON-309 URL: https://issues.apache.org/jira/browse/DAEMON-309 Project: Commons Daemon Issue Type: Bug Components: Procrun Reporter: Joachim Sauer Priority: Minor
The documentation for how the start method should act in JVM mode is conflicting (http://commons.apache.org/proper/commons-daemon/procrun.html). 1. In the documentation of the StartMethod parameter there's this text: {quote} Note: in jvm mode, the start method should not return until the stop method has been called. {quote} 2. The second-to-last sentence in the section "Using Procrun in jvm mode" reads: {quote} Note that the method handling service start should create and start a separate thread to carry out the processing, and then return. {quote} According to my reading those two sentences say pretty much the opposite of each other and observation shows that #1 is the correct one (i.e. the service is assumed to have stopped when the start method returns). Generally speaking, JVM mode is pretty under-documented (in my opinion). That can easily be verified by the fact that very popular documentation exists outside the procrun/commons daemon project, as a blog post: http://joerglenhard.wordpress.com/2012/05/29/build-windows-service-from-java-application-with-procrun/ A simple end-to-end example showing a similar setup as that blog post would be good. -- This message was sent by Atlassian JIRA (v6.1#6144)