Re: Pizza Example

In message <9Jugc.75998$vO4.5098632_at_phobos.telenet-ops.be>, Jan Hidders <jan.hidders_at_REMOVETHIS.pandora.be> writes
>> Given a known primary key, it takes on average about 1.05 disk seeks
>>(excluding OS file management overhead) to retrieve a record. Given a
>>known index value, it takes a further 1.05 disk seeks to retrieve a
>>list of ALL primary keys associated with that value. Added to that, a
>>Pick record usually contains all data associated with an entity -
>>that would likely be scattered across multiple rows in multiple tables
>>requiring multiple seeks in a relational db.
>
>Sure. Retrieving records with an index is also very fast with DBM
>files. But what happens if you have a query with a couple of joins, a
>little aggregation and some subqueries?

Apply a little statistics. How likely is that going to occur with an RDBMS, as opposed to a Pick DB. If the stats say it's unlikely to happen, then why worry about it?

That's what I meant about Pick queries being impossible to optimise - the statistics say there's no room for improvement, or that any gains by optimising the unusual case will be lost by hindering the typical case.

Cheers,
Wol

-- 
Anthony W. Youngman - wol at thewolery dot demon dot co dot uk
HEX wondered how much he should tell the Wizards. He felt it would not be a
good idea to burden them with too much input. Hex always thought of his reports
as Lies-to-People.
The Science of Discworld : (c) Terry Pratchett 1999