[postgis-users] Substantial different index use between 9.5 and 9.6 using Postgis

[Sandro Santilli]

On Sun, Dec 04, 2016 at 02:46:03PM +1100, Bill Measday wrote:
> This looks odd. Here is what pgAdmin tells me if is the current definition
> of st_setsrid (cost of 1.0)
> 
> CREATE OR REPLACE FUNCTION public.st_setsrid(
>     geog geography,
>     srid integer)
>     RETURNS geography
>     LANGUAGE 'c'
>     COST 1.0
>     IMMUTABLE NOT LEAKPROOF STRICT
> AS '$libdir/postgis-2.3', 'LWGEOM_set_srid'
> ;
> 
> However using you new definition still seems to have fixed the issue.

The one you just showed is for *geography*, while Regina fixed
the one for *geometry*.

--strk;