I moved forward with my project in Kotlin but, due...

# kotlin-spark

[Mike Lin]

I moved forward with my project in Kotlin but, due to the apparent maintenance risk with kotlin-spark-api, I regrettably chose to use Spark's Java DataFrame API (with 'manually' formulated RowEncoders) instead of relying on the Dataset type safety provided by kotlin-spark-api. This minimizes our dependence on the latter.