Robert Haas writes:
> The bit about maybe not getting both t and f as MCVs on a Boolean does
> seem a little worrying, but I'm not sure whether it actually affects
> query planning in a materially negative way. Can you demonstrate a
> case where it matters?
If we were trying to force that to hap
On Sat, Mar 24, 2012 at 12:17 AM, Dan McGee wrote:
> This is a bit of a corner case in all honesty, but if you have a short
> table (under 20 rows), the 10% heuristic used that decides whether
> distinct values scale with the row count will result in rather odd
> values for stadistinct in pg_stati
This is a bit of a corner case in all honesty, but if you have a short
table (under 20 rows), the 10% heuristic used that decides whether
distinct values scale with the row count will result in rather odd
values for stadistinct in pg_statistic, such as '-0.2' or '-0.67',
rather than the expecte
