Web Apps continued

I decided to check out the original post by Olivier Travers. Using the post and comments you get the following breakdown for clients, which use online internet services:
• desktop app - this is an application that runs on your client platform (PC, Mac, whatever). In order to use such an application you need to download the application and install it on your client platform. These applications are without an online services component. An alternative word for this is “standalone app”;
• internet app - this is a desktop app which has an online services element. Scoble calls these “non-browser apps or smart clients”;
• web app - applications that use a web-browser with embedded emulators (Flash, Javascript…) to access online services. No installation of the application is required. The use of these emulators allow for nicer interfaces. The word rich interface applications (RIA) is often used here. For the user this might imply that they have to download the (latest) version of the emulator. It is also argues that a web app might not be available on all platforms due to a lack of emulators on those platforms. Scoble calls these ‘Internet connected apps”. Note that from a user point of view, these applications are just “the web” as they use their Internet browser to access these apps;
• (pure) web app - applications that do not use emulators or interpreters to access online services. I guess that this implies only HTML and CSS. This category only uses protocols and markup which is standardized. One important point Olivier Travers makes is a Flash application does not expose its state to the browser. For him you need to support REST to be a web app. He also says that all content must be visible in order to be a (pure) web app. This implies that each piece of content must have an URL.

I am not sure how this breakdown works for other platforms, such as mobiles. I guess however that it will follow a similar pattern.