Re: BUG #13908: Query returns too few rows

On Thu, Feb 4, 2016 at 3:13 PM, David G. Johnston <
david(dot)g(dot)johnston(at)gmail(dot)com> wrote:

> Tom, we really need you to chime in here.
>
> On Thu, Feb 4, 2016 at 2:58 PM, David G. Johnston <
> david(dot)g(dot)johnston(at)gmail(dot)com> wrote:
>
>> Adding back the list so that everyone can see the latest finding. Adding
>> in the original bug block too...
>> ​
>>
>> The following bug has been logged on the website:
>>
>> Bug reference: 13908
>> Logged by: Seth
>> Email address: seth-p(at)outlook(dot)com
>> PostgreSQL version: 9.5.0
>> Operating system: Windows 10
>> Description:
>>
>>
>> * -> Unique (cost=1497322.60..1501340.65 rows=267870 width=24) (actual
>> time=153375.583..154829.461 rows=415874 loops=1)*
>>
>>
>
> ​Without running any DML I just got this result on the DISTINCT query...​
>
> *​ -> Unique (cost=1519634.64..1520973.99 rows=200 width=48) (actual
> time=161695.425..163174.422 rows=416075 loops=1)*
> ​
>

​According to 9.3 this 416,075 count is the correct one - and I got the
same count with and without distinct executed multiple times each.

​My next thought is to see which 300 or so rows that are being left out in
9.5 appear in 9.3....​

​David J.​