On the Stoimen.com blog there's a new post showing how you can connect your script to YouTube's API via the Zend_Gdata_YouTube component complete with pagination support.
The task is to read the entire Gdata from a YouTube's channel. It may sound easy, but as you may know Gdata is based on the Atom publishing protocol, and it naturally gives you only the latest few items.
The component lets you go through the pages of content from a user or channel on YouTube and pull out the metadata about each. Code is included to show you how to make the connection, get the thumbnail information for each video and push it back out the other side in an nice, clean array.