I just struck this too. liftmodules.omniauth pulls in gae-stuff that
breaks mailing in lift:
+-net.liftmodules:omniauth_2.10:2.5-RC2-0.7 [S]
[info] | +-net.databinder:dispatch-core_2.10:0.8.9 [S]
[info] | | +-org.apache.httpcomponents:httpclient:4.1.3
[info] | | +-commons-codec:commons-codec:1.4 (evicted by: 1.6)
[info] | | +-commons-logging:commons-logging:1.1.1
[info] | | +-org.apache.httpcomponents:httpcore:4.1.4
[info] | |
[info] | +-net.databinder:dispatch-gae_2.10:0.8.9 [S]
[info] | | +-com.google.appengine:appengine-api-1.0-sdk:1.5.5
[info] | | +-net.databinder:dispatch-http_2.10:0.8.9 [S]
[info] | | | +-net.databinder:dispatch-core_2.10:0.8.9 [S]
[info] | | | | +-org.apache.httpcomponents:httpclient:4.1.3
[info] | | | | +-commons-codec:commons-codec:1.4 (evicted by: 1.6)
[info] | | | | +-commons-logging:commons-logging:1.1.1
[info] | | | | +-org.apache.httpcomponents:httpcore:4.1.4
[info] | | | |
[info] | | | +-net.databinder:dispatch-futures_2.10:0.8.9 [S]
[info] | | | | +-org.scala-lang:scala-actors:2.10.0 [S]
[info] | | | |
[info] | | | +-org.apache.httpcomponents:httpclient:4.1.3
[info] | | | +-commons-codec:commons-codec:1.4 (evicted by: 1.6)
[info] | | | +-commons-logging:commons-logging:1.1.1
[info] | | | +-org.apache.httpcomponents:httpcore:4.1.4
[info] | | |
[info] | | +-org.apache.httpcomponents:httpclient:4.1.3
[info] | | +-commons-codec:commons-codec:1.4 (evicted by: 1.6)
[info] | | +-commons-logging:commons-logging:1.1.1
[info] | | +-org.apache.httpcomponents:httpcore:4.1.4
My quickfix is to just exclude the transitive gae dep:
"net.liftmodules" %% "omniauth" % (liftModulesLiftVersion+"-0.7")
exclude("net.databinder", "dispatch-gae_2.10"),
However, this sucks since this will break RUNTIME when I update to
scala 2.11. I couldn't get either:
"net.liftmodules" %% "omniauth" % (liftModulesLiftVersion+"-0.7")
exclude("net.databinder", "dispatch-gae")
nor ExcludeAll(organization="net.databinder" name="dispatch-gae")
to remove the dependency. Any tips on how to do this better? Couldn't
find any doc about wildcard exclusion in sbt…?
And isn't this really a bug that should be fixed in lift-omniauth?
Thanks,
Viktor