Github user tmalaska commented on the pull request:
https://github.com/apache/spark/pull/507#issuecomment-41319131
@pwendell to be honest this is a little deeper then I normally go with Pom
specification.
I think we are there is a behavior that maven does and there is a behavior
that sbt does and they are not the same.
My goal of this pull request was to make the same outcome for both maven
and sbt. In the end thrift will be included because would have Flume pulled it
in.
It's not a perfect solution but the only option is to change the pom in
Flume 1.4.0 but they have a different requirement of having two thrift options.
I'm not even sure how Flume would honor that requirement without profiles.
Also I figured it would be good to have Flume 1.4.0 in Spark 1.0, because
Flume 1.4.0 is the most commonly used Flume out there and it has some really
cool functionality I would like to add to the FlumeStream like compression and
encryption.
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at [email protected] or file a JIRA ticket
with INFRA.
---
