David wrote:

I'd like to call a vote for releasing the axistools 1.2 plugin.

+0, not using it.

There are issues left - and some cleaning up could still be a really good idea,

Just to make sure, it's explicitly mentioned:

"mvn docck:check" reports quite a dozen of mojo parameters which have no documentation.

Also, the POM is missing a <license> element (seems to be ASL 2.0).

"mvn dependency:analyze" shows some oddities with the Maven dependencies.

The site descriptor can be trimmed down, e.g. banners and links are inherited.

The POM snippets on the plugin's usage/example pages don't include a <version> element as suggested by best practices.

Furthermore, the POM snippet on the usage page does not even give a <groupId> which won't work since the default value is "org.apache.maven.plugins".

The mojo parameters
- project in AdminMojo
- project in Java2WSDLMojo
- project in WSDL2JavaMojo
- pluginArtifacts in WSDL2JavaMojo
should be marked as @readonly to prevent user confusion.

The mojo parameter
- timestampDirectory in WSDL2JavaMojo
defaults to "${basedir}/target" which is better described by "${project.build.directory}".

Likewise, the mojo parameter
- classesDirectory in Java2WSDLMojo
defaults to "${project.build.directory}/classes" which matches "${project.build.outputDirectory}".

The method getRuntimeClasspath() in Java2WSDLMojo looks strange. First, it neglects class path ordering as specified in the POM by storing the paths in a HashSet. Second, the method queries getRuntimeClasspathElements() although the mojo is only annotated with "@requiresDependencyResolution compile", i.e. the plugin will usually not see anything beyond compile scope.

The mojo parameter
- staleMillis in WSDL2JavaMojo
is never read (c.f. PMD report). Looks like a missing call to WSDL2JavaPlugin.setStaleMillis().


Benjamin

---------------------------------------------------------------------
To unsubscribe from this list, please visit:

   http://xircles.codehaus.org/manage_email


Reply via email to