[javadocs] Fix javadocs

[hazendaz]

option needed to have additional arguments, removed old source 8 on javadocs as no longer needed here.

• Replacement of Fix javadoc options / Use legacy mode to overcome <source> level issue #3231

[scordio]

main is not yet open for contributions, please target 3.x for now (where <additionalOption> is already used).

What I meant in the other ticket is that I don't get the problem we're trying to solve with the introduction of:

<legacyMode>true</legacyMode>

As far as I can judge from the logs and the output at https://www.javadoc.io/doc/org.assertj/assertj-core/latest/index.html, we don't have any visible issue.

In addition, compared to the JUnit Javadoc, we don't have the MODULE entry in the top header. I honestly don't know why. Could it be because we currently compile module-info.java in a separate maven-compiler-plugin execution?

assertj/assertj-core/pom.xml

Lines 321 to 333 in 4f072c9

	<execution>
	<id>jdk9</id>
	<goals>
	<goal>compile</goal>
	</goals>
	<configuration>
	<release>9</release>
	<compileSourceRoots>
	<compileSourceRoot>${project.basedir}/src/main/java9</compileSourceRoot>
	</compileSourceRoots>
	<multiReleaseOutput>true</multiReleaseOutput>
	</configuration>
	</execution>

Could it be related to the multiReleaseOutput parameter which will put the module-info.class under META-INF/versions/9?

Honestly, I didn't dig into the maven-javadoc-plugin to cross-check the sources used.

[hazendaz]

Ah, I see. Yes because you have module-info file it doesn't require that now. The bug maven had was with automatic module naming. I can close this then and reopen to the other branch with just the expected content here without that part.

[hazendaz]

3.x line already has the change so I guess nothing else to do here