Re: How Postgresql.... Putting Count(*) To Bed (Hopefully)

On Sunday 22 July 2001 15:49, you wrote:
> Mark kirkwood <markir(at)slingshot(dot)co(dot)nz> writes:
> > I have not tried to see if it makes any difference in Postgresql...
>
> Postgres converts count(*) to count(1) at the grammar stage (cf. gram.y,
> about line 4817 in current sources). So if you think you detect any
> performance difference, you're surely hallucinating...
>
> regards, tom lane

It seemed appropriate to test my "experimental method" with this example :
I measured elapsed times for count(*), count(1) ....

I can (with some relief) report elapsed times for both at 1m01, with a
variation of 1 s for both measurements ...

thanks for everyboys patience here...

Mark