There are no fair benchmarks between PostgreSQL and MySQL, and that's because MySQL's functionality is too limited to do any fair benchmarks.
Simple example:
SELECT name FROM table WHERE LOWER(name)='bert';
would cause a sequencial scan on MySQL because MySQL cannot create an index that contains 'LOWER(name)', but in PostgreSQL you'd define a functional index on 'LOWER(name)', et presto; postgreSQL wins.
MySQL's strong point is very high speed at simply queries on small tables at few concurrent users.
...and that is exactly all that MySQL is good at.
MySQL does not support subqueries, functional indexes, stored procedures, cascading delete/update on PK/FK relations, no functions like INTERSECT and EXCEPT, and no VIEWs.
So basically MYSQL lacks all the features that make a database usefull :-)
MYSQL also has a very annoying 'feature' of being able to swich databases once a connectin has been made, wich sounds spiffy but has f*cked up more than one application.
The performance of a database is not just speed, it is also functionality and reliability.
Something as simple as cascading delete on a PK/FK relationship can really mess up your application, because you have to check for orphan records manually every single time you delete any record that can be refered to anywhere in the database. That can quickly mean running 5-6 queries to delete one record.
In postgresql, it's just one query and you can trust the database to remove all related data quickly and reliably.
A forum, a FAQ, what else do you need?
