I don't see any significant code change (
https://github.com/gwtplus/google-gin/compare/v3.0.0...2073039; besides reformatting, removing redundant 'public' modifiers, small javadoc tweaks, and a method renamed in tests), so is it even worth cutting a new release?
I mean, people will (should!) have the GWT and Guice dependencies already (and in Maven, it will in most cases override the transitive versions from GIN), so given the absence of code change, they should be able to use 3.0.0 with GWT 2.9.0 and Guice 4.2.3 already (and use HEAD-SNAPSHOT with GWT 2.8.2 and Guice 4.2.2; or any combination).