I think the moment we start touching catalyst we should be using Scala. If
in the future there is a stored procedure api in Spark we can always go
back to Java.
On Tue, Aug 25, 2020, 4:59 PM Anton Okolnychyi
wrote:
> One more point we should clarify before implementing: where will the SQL
> exte
One more point we should clarify before implementing: where will the SQL
extensions live? In case of Presto, the extensions will be exposed as proper
stored procedures and can be part of the Presto repo. In case of Spark, we
could either keep them in a new module in Iceberg or in a completely di
