Whoa, hoss! Let's not start calling web apps "rich clients" just yet
One thing I've been hearing a lot lately is the misuse of the term "rich client" when describing a web application heavily using client-side functionality or aesthetic enhancements (I've incorrectly uttered the phrase a couple of times myself). This is most common in web creations having undergone hardcore Dynamic HTML treatment and/or exhibiting AJAX-style programming.
This bothers me because the term "rich client" still typically refers to a desktop Windows program enhanced with sophisticated functionality that's largely only emulated - not perfectly replicated - on the Web. Examples of such are WinForms with features like drag-and-drop, auto-complete, complex data persistence, embedded databases, fancy real-time visual effects, or the ability to instantly drill-down into a complex list. All typically don't carry the additional baggage of server roundtripping, which is costly in terms of both back-end performance and front-end experience.
To say that a web client is "rich" even in the most elaborate settings such advanced behavior is inaccurate.
Face the music, people: the functional divide between web apps and their desktop equivalents (the latter of which is always credited with being the ultimate in terms of creating a winning user experience) is a heckuva lot more narrow than in years past, and with things like AJAX, XML web services, globalization, and other features it's starting to get even narrower. We may even reach the point someday where an application can be as thin as one that's based on the web and as easily deployable, but as thick and exhibiting as much remote server autonomy as a desktop app.
Just not yet. So let's all be patient, our terms collectively straight, and call a spade a spade.