It's not exactly true that "calcite came from drill". Calcite was originally called Optiq. Drill was the second project to use Optiq (Cascading was the first) and Optiq was a pretty significant code base (almost 200k lines of code) when Drill started to use it.
Drill created their own branch/fork of Calcite but didn't keep up with Calcite master[1], and eventually they gave up. (Blame on both sides. It's hard to evolve a framework because people are using APIs, and relying on behavior, that you never thought of as "public". So, each time Calcite releases it is a small tax on Drill to keep up.) Much of the work done to Calcite applies to the core (e.g. better rules for pushing filter through grouping sets, improving how we deduce constraints and simplify expressions) and Drill could benefit from that. Drill do cherry-pick the occasional feature. Julian [1] https://issues.apache.org/jira/browse/DRILL-4093 On Sun, May 28, 2017 at 5:49 AM, Ted Dunning <[email protected]> wrote: > That is a good question. That basic answer is probably a measure of the > success of calcite and also the fact that calcite came from drill so most > of the changes since calcite was broken out have been to meet the needs of > other consuming projects. > > Regardless, it would be nice to keep up. > > > > On May 28, 2017 1:43 PM, "Muhammad Gelbana" <[email protected]> wrote: > > Drill is using Calcite v1.4 while it has already reached v1.12. For myself, > there is no reason to upgrade at the moment but I'm wondering why haven't > Drill upgraded it's used version of such a core component as Calcite for a > bit more than a year and a half now ? > > *---------------------* > *Muhammad Gelbana* > http://www.linkedin.com/in/mgelbana
