Github user kiszk commented on the issue: https://github.com/apache/spark/pull/19601 I agree with you that we need to improve the write path. It will be addressed after improving the frequently-executed read path, as you suggested before. It will be addressed by the following PR. For improving the read path, which approach is better? To add new `ColumnVector.Array` or to add new `WritableColumnVector`?
--- --------------------------------------------------------------------- To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org For additional commands, e-mail: reviews-h...@spark.apache.org