A while back, DevShed.com posted Part 1 of an ADODB tutorial, showing you how to use the ADODB abstraction layer in your scripts to make them more portable from platform to platform, machine to machine.
Well, this morning, they've posted the second part in the series - PHP Application Development With ADODB (part 2). This one skips all the introduction kinds of things and gets straight to the "meat" of the abstraction layer. They talk about rapid execution of similar queries, commits, cached queries, and exporting your results, all using built-in functions that ADODB already has. I have to admit, using the cached and similar queries functions would of been a life saver for me several times over. I can think of lots of times that I was just running the same selects over and over again. Cached results take tons of the load off the site and make life easier for both you and your site (and trust me, your site will thank you for it!).