st 6. 10. 2021 v 5:50 odesílatel napsal:
> Hello,
>
> there is a bug with types in functions. Here is an example:
>
> --drop type xyz;
> create type xyz as ( x numeric, y numeric, z numeric );
>
> --drop table test_xyz ;
> create table test_xyz ( a int, b xyz, c xyz );
> insert into test_x
On Tuesday, October 5, 2021, wrote:
>
> create function test_xyz3() returns table ( b xyz) as
> $$
> declare
> bb xyz;
> cc xyz;
> begin
> select b, c into bb, cc from test_xyz ;
> return bb ;
> end;
> $$ language plpgsql ;
> -- --
>
> select * from test_xyz3() ; -- compiling error
>
On Tuesday, October 5, 2021, wrote:
>
> select * from test_xyz2() ; -- 3 columns, but 1 should be correct?
> select b from test_xyz2() ; -- error
> select x from test_xyz2() ; -- never declared column
All three of these stem from:
https://www.postgresql.org/docs/current/xfunc-sql.html#XFUNC-SQ
On Wed, 2021-10-06 at 00:21 -0500, Ron wrote:
>
> What version of Postgresql do you think has the bug?
> Where did you get the Postgresql binaries from?
> What platform?
FreeBSD
pkg install postgresql ...
Version 13.3
a friend use Debian unstable with PG version 14 and have the same
problem.
>
What's the definition for abc_serv_nch_q1_2021_expr_idx3? That is a jsonb
field I assume? Statistics aren't great on jsonb data, so you may benefit
greatly from pulling keys out to be stored as a standard column. I would be
curious for more "pure" estimates on each quarterly partition directly for
What version of Postgresql do you think has the bug?
Where did you get the Postgresql binaries from?
What platform?
On 10/5/21 10:50 PM, m...@ft-c.de wrote:
Hello,
there is a bug with types in functions. Here is an example:
--drop type xyz;
create type xyz as ( x numeric, y numeric, z num
Hello,
there is a bug with types in functions. Here is an example:
--drop type xyz;
create type xyz as ( x numeric, y numeric, z numeric );
--drop table test_xyz ;
create table test_xyz ( a int, b xyz, c xyz );
insert into test_xyz values
( 1,(11.1, 12.3, 14.5 ), (22.2, 22.3, 24.5 ) ) ;
Just a quick update to this topic from my testing: I whipped up a quick python
script to create the partition tables for me, which went smoothly enough, and
created a table LIST partitioned on station. Once populated with my current
data, this table proved marginally faster than the unpartitione
Thanks Ben,
How is life treating the AusVet these days? Been quite a few versions under the
bridge since we discussed R and Postgres stuff. 2009 was it??
That kicked some quiescent brain cells into gear. I just tried it from home,
where it also works fine, if a bit sluggish, which seems to be h
Hi Brent,
not sure why, but it is working for me using ogrinfo:
~ $ ogrinfo
https://dservices3.arcgis.com/fp1tibNcN9mbExhG/arcgis/services/BOMEC_15_Class_WFS/WFSServer\?SERVICE\=WFS\&REQUEST\=GetCapabilities
INFO: Open of `
https://dservices3.arcgis.com/fp1tibNcN9mbExhG/arcgis/services/BOMEC_15_C
10 matches
Mail list logo
