Re: integer ceiling in LIMIT and OFFSET - Mailing list pgsql-hackers

From Rod Taylor
Subject Re: integer ceiling in LIMIT and OFFSET
Date
Msg-id 1066843522.3284.27.camel@jester
Whole thread Raw
In response to Re: integer ceiling in LIMIT and OFFSET  ("scott.marlowe" <scott.marlowe@ihs.com>)
Responses Re: integer ceiling in LIMIT and OFFSET
List pgsql-hackers
On Wed, 2003-10-22 at 12:08, scott.marlowe wrote:
> On Wed, 22 Oct 2003, Tom Lane wrote:
>
> > Rod Taylor <rbt@rbt.ca> writes:
> > > That said, perhaps the TODO for changing LIMIT / OFFSET to be expression
> > > based should also mention bumping them to int8.
> >
> > Can't get excited about it ... this would slow down the normal use of
> > the facility for what seems a completely hypothetical need.
>
> While I'm pretty sure
>
> select * from sometable limit 2147483648
>
> isn't gonna be common, maybe someone would be likely to do something like:
>
> select * from sometable limit 10 offset 2147483648

> I wouldn't do it, but who knows what shadows lurk in men's minds?

The overhead in simply getting to that offset is going to be significant
and I would place my bets against anyone attempting that.

A cursor pulling small chunks of a multi-billion tuple set is probably
much more common, so we should ensure those work.



pgsql-hackers by date:

Previous
From: Bruce Momjian
Date:
Subject: Re: So, are we going to bump catversion for beta5, or not?
Next
From: Tom Lane
Date:
Subject: Re: 7.4 compatibility question