Re: WIP patch: distinguish selectivity of < from <= and > from >=

Ashutosh Bapat <ashutosh(dot)bapat(at)enterprisedb(dot)com> writes:
> On Tue, Jul 4, 2017 at 9:23 AM, Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> wrote:
>> regression=# explain analyze select * from tenk1 where thousand < 10;
>> before:
>> Bitmap Heap Scan on tenk1 (cost=5.14..241.38 rows=110 width=244) (actual time=0.121..0.623 rows=100 loops=1)
>> with patch:
>> Bitmap Heap Scan on tenk1 (cost=5.06..227.42 rows=100 width=244) (actual time=0.054..0.300 rows=100 loops=1)

> It's expected that the estimates will change with this patch. But I am
> wondering why should actual times vary so much. May be that's just
> accidental. Butthe actual timings are consistently lower with the
> patch except the last one

I didn't intend those examples to illustrate anything except row counts,
so I didn't worry about making the timings comparable. Some of the
examples were probably the first query in a session and so would've faced
cache-populating costs the others didn't. Also, the "before" examples
were all done on 9.6 not HEAD.

regards, tom lane