If we ever fix test-jar packaging the optional will make sense ;-) On Tuesday, 14 August 2012, Laird Nelson wrote:
> One other "I've spent too much time on this" question. > > I've been working with Maven for years, so understand it pretty well. But > this particular combination of options has always stumped me. > > I've read > > http://maven.apache.org/guides/introduction/introduction-to-optional-and-excludes-dependencies.html > , > and I've read > > http://maven.apache.org/guides/introduction/introduction-to-dependency-mechanism.html#Dependency_Scope > . > > Suppose I have a dependency in test scope. OK, I get that. It will be > included on my test classpath. > > Now suppose someone depends on me. First suppose they depend on me in > compile scope. I know that my test dependency does not end up on their > compile classpath. But does my test dependency get downloaded to their > local repository? > > If it does not, is there any sense in having a dependency declared like > this: > > <dependency> > <groupId>whatever</groupId> > <artifactId>whatever</artifactId> > <version>whatever</version> > <scope>*test*</scope> > <optional>*true*</optional> > </dependency> > > > ...? > > Another way to ask this: doesn't test scope sort of imply "optional"? (If > the answer is "no", I'd appreciate some kind of reasoning if it's not too > much trouble.) > > Best, > Laird > > -- > http://about.me/lairdnelson >
