> I don't see how it could be:
>SELECT * FROM (SELECT * FROM a ORDER BY x DESC) AS B ORDER BY x ASC;
That is a rather different query. My question was if the order of two
elements whose internal order is not affected by the current ordering
clause, still may change places due to technicalitie
Is sorting in PostgreSQL stable over subqueries, that is, is
select * from (select * from A order by x) as B order by y;
equivalent with
select * from A order by y, x;
?
---(end of broadcast)---
TIP 2: Don't 'kill -9' the postmaster
> Looks like something for ltree, which is a datatype for storing
> tree-structured data.
>
> http://projects.commandprompt.com/public/pgsql/browser/trunk/pgsql/contrib/ltree/README.ltree
Unfourtunately, ltree seems to compare node labels in lexicographic
order:
"have their usual meanings. Compa
I need to save something very similar to X.50x IOD:s, that is, paths
where the path-components are numbers. For example: 10.5.5003.24.35.
Futhermore, I need to sort these in numerical path order, so that if
two paths are compared according to the first path component that
differs, and this path co
If you do not want to amend your table with extra information, this is
how you do it:
Suppose you have a table
create table events (
time timestamp,
object int refers objects(id), -- The thing that had its ignition
turned on or off at this time
ignition boolean,
comment varchar
);
You can th
> Nope, the query is way too expensive to run it twice. (GIN scan over
> well over half a million rows. It's faster to do the
> get-as-cursor-then-loop-and-increment, I've measured that)
Can't you select into a temporary table and then do select count(*)
over that table, aswell as run your cursor
> select count(*) from new ??? - new refers to a record not a table name
In the description in the manual, new and old are described as
relations... Also, isn't the point that the query is rewritten into the
new query with rules, thus if you update more than one row, new should
be all the updated
