I can't tell you much about mysql, but I have been using postgresql for a
few years now. On the last project I was on, we were using postgres to
handle a database that grew by more than 1GB/day (I don't recall how many
millions of rows that represented), and we were able to tune it to get the
performance we needed on pretty standard servers. It supports
transactions, triggers, stored procedures, subqueries, and I have yet to
come up with a join that it couldn't do (In my current project, we
routinely use every sort of join except for a cross join to get our data
out the way we like it). If you're doing anything the IPs or network
addresses, it also has some sweet special data types and operators for
them.
As for an unbiased comparison between postgres & mysql, your best bet
might be to talk to people who are good at one or the other and make your
own comparison... Generally speaking, the people with the knowledge you
seek are already fairly biased towards the one they use...
--Nick
(text/plain)