I looked at https://maven.apache.org/plugin-tools/apidocs/org/apache/maven/plugins/annotations/Parameter.html#readonly() and my interpretation of this annotation parameter is that it exist for a use case it don't fully understand (usage of common POM elements).

My guess is that this readonly parameter is not related to Java attribute immutability. I'm also (mostly guessing) that the original author (@olamy ?) copy-pasted from the "wrong" thing. Perhaps from MavenProject/Settings/MavenSession? My assumption is that the intention is that the end-user should be able to configure this parameter in a <configuration> block. While reading the javadoc for readonly it looks like this should be false (or not defined, so it is false per default) for the indented use case of specifiying certchain.

I have rebased this pull request on master.

readonly is not about documentation. See https://maven.apache.org/plugin-tools/maven-plugin-annotations/apidocs/org/apache/maven/plugins/annotations/Parameter.html#readonly()

It's possible that this should not have readonly=true. I don't have a real opinion on that. But let's make sure we're making conscious choice here and the PR title reflects what we're trying to do.

Even though readonly is not about documentation (primary), I feel that in this context (for this specific case and JIRA tickets) it is about documentation. In my comment in https://issues.apache.org/jira/projects/MJARSIGNER/issues/MJARSIGNER-63 I have provided a (rather long) example that shows that it is possible to set this parameter, even if it is readonly (or undocumented).

My opinion is that we (the community) should make a conscious choice to make sure that this parameter is configurable by the end-user (and also documented). I think this was the original intent of https://issues.apache.org/jira/projects/MJARSIGNER/issues/MJARSIGNER-53

Resolve #108