I've updated RssFeed, the open-source custom ASP.NET server control for displaying RSS feeds in an ASP.NET Web page. The control is now at version 1.3. The latest update doesn't change any functionality of the control, it just gives it a slight performance kick in the pants. Specifically, previous versions loaded the RSS feed into an XmlDocument and then used XPath expressions to get the <item> elements, and then more XPath expressions to get the <title>, <author>, <description>, <category>, and <author> elements out of the current <item>.
I replaced this code with the more performant XPathDocument / XPathNavigator / XPathNodeIterator, since I was just doing XPath expressions and read-only operations. The move was made per suggestion of Aaron Skonnard's article on .NET XML Best Practices. After making the change I ran some rather unscientific bench mark tests using The Microsoft Web Application Stress Tool (WAST), and found for light loads the XmlDocument approach seemed as efficient as the XPathDocument approach, but under heavier loads, the XPathDocument approach was slightly more efficient (but nothing drastic). I had read in others' blogs that XPathDocument was way more efficient than XmlDocument - perhaps I didn't see extreme performance gains due to the relatively small size of the typical RSS file.