On 15/04/2019 22:32, Peter Geoghegan wrote:
> On Mon, Apr 15, 2019 at 12:19 PM Tom Lane <tgl@sss.pgh.pa.us> wrote:
>> Perhaps, but we won't know if we don't try. I think we should try,
>> and be willing to add hooks and flexibility to core as needed to make
>> it possible.
>
> We could approach it without taking a firm position on inclusion in
> core until the project begins to mature. I have little faith in our
> ability to predict which approach will be the least painful at this
> early stage.
When we started hacking on this, we went in with the assumption that
this would have to be in core, because WAL-logging, and also because a
column-store will probably need some changes to the planner and executor
to make it shine. And also because a lot of people would like to have a
column store in PostgreSQL (although a "column store" could mean many
different things with different tradeoffs). But if we just have all the
necessary hooks in core, sure, this could be an extension, too.
But as you said, we don't need to decide that yet. Let's wait and see,
as this matures.
- Heikki