Github user JoshRosen commented on the pull request:
https://github.com/apache/spark/pull/1639#issuecomment-50576330
This method is intended to be called by Scala classes that implement
Java-friendly wrappers for the Spark Scala API. For instance, MLlib has APIs
that accept RDD[LabelledPoint]. Ideally, the Java wrapper code can simply call
the underlying Scala methods without having to worry about how they're
implemented. Therefore, I think we should prefer the `retag()`-based approach,
since `collectSeq` would require us to modify the Scala consumer of the RDD.
Since this is a private, internal API, we should be able to revisit this
decision if we change our minds later.
---
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.
---
