Or "still picking on Roger" - specifically this part of the comment quoted yesterday:
There are many free databases to choose from: Firebird, Ingres, Postgres, and MySQL to name (most of) them. Most of the "best" developers and the largest community are behind Postgres. MySQL came out of nowhere - and it seems destined to return there, after the can-do founders left the horribly middle-managed Sun Microsystems.
I recently did some rough tests related to this - setting up Drupal against both PostgresSQL and MySQL, handing them both about 25 million docs, disconnecting the power to various system components to check recovery, and hitting both with purposefully erroneous PHP and PerlDBI scripts. Neither showed any serious problems and the difference really was that MySQL was significantly lighter across the board - meaning it used less of my time, less RAM, and fewer processor cycles.
I don't want to predict where MySQL is going in the long run, but right now it seems fast, simple, reliable, and generally adequate to the purpose.
PostGresSQL did the job too - but it's (despite the SQL part) vastly more advanced in terms of data generality and correspondingly overkill for this job: it feels like using a highway bus to go grocery shopping.
So MySQL works for this job and Roger's second question now is: do I care that the MySQL founders have left Sun and lots of people are becoming less enthused about contributing to its development?
I don't - because my primary concern is stability: the systems' ability to routinely execute routine operations. There's nothing fancy needed here: it does what it does, and as long as it continues doing that, having the founders move off is actually good news: not only is there reason to hope one or two of them will go on to create something else that's new and exciting, but the truth is that genius creates advances but also bugs and destabilization where what I want is for the thing to run unchanged for many years.