Re: A question for Mr. Celko

From: Jan Hidders <jan.hidders_at_REMOVETHIS.pandora.be>
Date: Mon, 19 Jul 2004 09:34:51 GMT
Message-Id: <pan.2004.07.19.09.35.30.941204@REMOVETHIS.pandora.be>

On Mon, 19 Jul 2004 04:08:48 +0000, Marshall Spight wrote:
> "Jan Hidders" <jan.hidders_at_REMOVETHIS.pandora.be> wrote in message
> news:pan.2004.07.19.00.19.32.644229_at_REMOVETHIS.pandora.be...

>>
>> That's a physical layer problem and the ideal database should let you
>> choose the right data structure for the mix of updates and queries that
>> you expect.

>
> (Actually, I would propose that the *ideal* database would measure the
> updates and queries it is receiving and choose the right data structure
> itself.)

Sure, but let's first try to take a few small steps before we see if we can run, ok? :-)

>> So the question is not "what is the best data structure" but "what data
>> structures are there" and in which case should we use which data
>> structure.

>
> Okay. So I'll ask: what data structures are there?

For lists? That's a whole research field in its own. You really expect me to give a short summary of that here? There are whole conferences dedicated to optimization in list-manipulating languages.

> And what kind of join executions are there? I guess we have nested loops
> and merge-joins. [...]

These are algorithms for *set* operations, not *list* operations. As I already said, yes, sometimes can be used for certain list operations, but the difficulty is to figure out when.

Jan Hidders

Received on Mon Jul 19 2004 - 04:34:51 CDT