Hi Peter,
On 10/21/21 22:16, Peter Geoghegan wrote:
I returned to the 1995 paper "Efficient Search of Multidimensional
B-Trees" [1] as part of the process of reviewing v39 of the skip scan
patch, which was posted back in May. It's a great paper, and anybody
involved in the skip scan effort should read it thoroughly (if they
haven't already). It's easy to see why people get excited about skip
scan [2]. But there is a bigger picture here.
Thanks for starting this thread !
The Index Skip Scan patch could affect a lot of areas, so keeping MDAM
in mind is definitely important.
However, I think the key part to progress on the "core" functionality
(B-tree related changes) is to get the planner functionality in place
first. Hopefully we can make progress on that during the November
CommitFest based on Andy's patch.
Best regards,
Jesper
