When programming an XI application, there will be three or more tool kits underneath your code on each machine you are supporting. For example, if you are writing an XI application for Microsoft Windows, you will encounter the XI tool kit for Windows, XVT for Windows and the Windows SDK. These tool kits are layered, with the “native” tool kit on the bottom of the stack, XVT for that platform on top of the native tool kit and XI on top of XVT.

When we say that a tool kit “sits on top” of another tool kit, such as XI on top of XVT, we mean that the tool kit on top uses the functions of the tool kit below it, but not vice-versa. XI can call XVT functions, but XVT doesn’t use any of the symbols defined by XI. In addition, we often say that a programmer can “drop down a level” to access the functionality of a tool kit beneath it. Since your code is on top of all of the tool kits in a layered API, you can access any layer anywhere in your program.

An XI interface is a window containing any number of user interface objects. Most XI objects are operable by the user and will invoke some action in the application. (Some XI objects are display-only.) Together the objects inside the interface behave in a coherent fashion to achieve the objectives of the program. You can think of these objects as analogous to controls in the XVT parlance or widgets in the vocabulary of some native tool kits.