Github user rxin commented on the pull request:
https://github.com/apache/spark/pull/3505#issuecomment-66411889
Let me explain - Spark SQL is more than SQL. It is SQL + a dsl that will be
improved over time. I personally believe over time, majority of Spark users
will interact directly with SchemaRDD instead, because that is optimized for
structured data. It is much easier in Spark SQL to also optimize for data
structure. Personally I'm worried about pushing more and more stuff into Spark
core itself because it is very hard to maintain and optimize for arbitrarily
structured JVM objects.
Most of the code you have written here can be used directly in SchemaRDD.
If we really want to apply this in core itself, even without considering the
maintenance burden, we will need to find a way so this can be turned on and off
easily (e.g. different method names) rather than relying on immutable confs.
---
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.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
