In general, Simulated objects are considered a black box where Rapise cannot make any firm assumptions. A good example is Canvas where the application draws something (it may be text, graphics etc). Rapise may learn how to reach the canvas object , but it is not able to identify the objects inside.

Canvas may be flexible, information in it may depend on system DPI, screen resolution etc. So there are two approaches for such situations:

  1. Try to look inside canvas. The canvas may be an applet exposing some UI Automation technology (such as flash etc) and there may be APIs to read it.
  2. Try to fix the environment. For instance, make application full screen before test execution. Make sure screen size is the same.

In addition to these 2 approaches there is plenty of possible workarounds available for specific situations. For example, sometimes we cannot grab information from an <img> element but we may take it from alt="..." attribute or from src.