revert some scaladoc changes because of scala 2.12 build issues

[pjfanning]

I've tested this and this revert allows me to publish the pekko-cluster-sharding-typed jar locally while it fails without this revert.

Reverts part of #682

[mdedetrich]

@pjfanning Wouldn't it be safer to use FQCN rather than just removing the prefix?

[pjfanning]

@pjfanning Wouldn't it be safer to use FQCN rather than just removing the prefix?

This is a partial revert of #682 because our builds are broken.

Scaladoc - especially ScalaDoc in Scala 2.12 has bugs. It is a minefield trying to get it right. My main objective is to fix the build. I can't get a build to test the sbt-assembly issue due to the broken build.

[mdedetrich]

This is a partial revert of #682 because our builds are broken.

Scaladoc - especially ScalaDoc in Scala 2.12 has bugs. It is a minefield trying to get it right. My main objective is to fix the build. I can't get a build to test the sbt-assembly issue due to the broken build.

I understand, but looking at the diff we are not changing from FQCN to a nested package prefix, rather we are changing from a nested packaged prefix to another nested package prefix.

I am quite sure that if you use a complete FQCN i.e. changing [[javadsl.Entity]] to [[org.apache.pekko.cluster.sharding.typed.javadsl.Entity]] (or w/e it *should * be) would actually remove all brittleness regarding the scaladoc classname lookup and help prevent future problems.

[pjfanning]

closing in favour of #688

[mdedetrich]

closing in favour of #688

@pjfanning I think you forgot to close?