Re: [Fwd: Re: no ORDER BY in subselects?]

At 11:50 21/09/00 -0400, Frank Bax wrote:
>At 11:29 AM 9/21/00 +1000, you wrote:
>>The main reason I use them is to find the 'next' or 'previous' record in a
>>list (eg. next date, next ID). eg.
>>
>> select <whatever>, (select ID from table where id > this.id
>> order by id asc limit 1) as next_id ...
>
>Doesn't this give the same result (without order by):
>
>> select <whatever>, (select min(ID) from table where id > this.id) as
>next_id

Yes, but I don't think PostgreSQL is smart enough to use indexes to
evaluate the Min() function.

Also, min/max does not work quite so well with a slightly more complex
example:

select <whatever>, (select ID from table where date_field > this.date_field
order by date_field asc limit 1) as next_id ...

(ie. if the date_field and id are not correlated, but you want the id
corresponding to the next date).

----------------------------------------------------------------
Philip Warner | __---_____
Albatross Consulting Pty. Ltd. |----/ - \
(A.B.N. 75 008 659 498) | /(@) ______---_
Tel: (+61) 0500 83 82 81 | _________ \
Fax: (+61) 0500 83 82 82 | ___________ |
Http://www.rhyme.com.au | / \|
| --________--
PGP key available upon request, | /
and from pgp5.ai.mit.edu:11371 |/