My understanding is that currently in order me to create date hierachies (e.g. year -> quarter -> month -> day, etc.) I would have to create a Hive table that contains the columns required to create such hierarchies and join that table to your date dimension attribute.
This is something that almost everyone using a tool like Kylin would probably want to do. So why not have that supported out of the box? Or is the preferred way to do this sort of thing at the query level: select year(dt), quarter(dt), count(*) from lookup_tbl group by year(dt), quarter(dt) -TPP
