When I haven't been working on getting this blog system written (more on that in another post), I've been playing with Yahoo's new search API. Yahoo has the Google-killing app here -- not only do they allow a sizable number of queries per day (5000, versus Google's 1000), it has quite a few features (video search, picture search, localized searches, not to mention your ordinary web search) and uses XML REST queries instead of SOAP. Developers seem to like SOAP, and for apps that need the security it's fine, but for ordinary results, you can't beat REST. It's what I use to get the Amazon info for DVD Verdict, and it's much easier to request something by URL, get an XML document back, and parse it appropriately than worry about encapsulating the query in SOAP, sending it as a POST request, getting back the SOAP-encapsulated XML, then parse it.

Where was I? Oh yeah, Yahoo. My first foray into their API was using the image search to write a spider to grab pictures. I used Perl's LWP module to download the picture, making sure I specified a non-obvious user agent and the referral URL as the request referer (since most sites don't allow off-site requests). Worked like a charm. With some more research, I should find a way to integrate it with DVD Verdict. Maybe something like this...