Re: Literature on physical algebra

Leandro Guimarães Faria Corsetti Dutra <leandrod_at_mac.com> writes:

> Kai Großjohann wrote:
>

>> For a database system, one needs a query language.  I have this.  One


> 

> 	I hope is a valid D, according to Date and Darwen.



Dunno.  I found http://citeseer.nj.nec.com/darwen95third.html -- is

>> also needs a logical algebra to provide a starting point for query
>> optimization.  I also have this.  But for actually doing query
>> optimization, one needs a physical algebra.  I still need the physical
>> algebra.


> 

> 	There's no such thing... algebra is a logical access and

Yes, but in addition to the access methods one needs operators to frob the results. I think this is also called a query execution plan. It would be nice to know equivalence relations¹ of different query execution plans, such that one could be sure to choose an efficient plan which produces the right result. No sense in choosing a fast plan which produces the wrong result...

I've discovered a book by Ullman which talks about the physical layer. I guess that's what I'm looking for. http://www-db.stanford.edu/~ullman/pub/dscbtoc.txt

kai

¹ In terms of results.

-- 
Linux provides a nice `poweroff' command, but where is `poweron'?