Automated benchmarking of software performance

A method for benchmarking a plurality of views of a window to a obtain computing performance measurement, the window having an associated view list comprising the views of the window, each view in the view list dividing the window by functionality and associated with a set of operations for altering a state of the window, the method comprising:

• selecting a plurality of views from the view list of the window for benchmarking;
• for each particular view of the plurality of selected views of the window:
• altering the state of the window by performing at least one operation from the set of operations associated with the particular view;
• and
• obtaining a computing performance measurement based on a time required to alter the state of the window, wherein the selecting, altering, obtaining, and benchmarking are performed automatically at a computer.

A non-transitory computer readable storage medium storing a computer program for benchmarking a computing performance of a window of an application running in a graphical user interface, the window having an associated view hierarchy comprising views of the window in parent and child relationships, the computer program comprising sets of instructions for:

• analyzing the view hierarchy of the window;
• based on the view hierarchy analysis, selecting a set of views of the window for benchmarking, each view describing a particular region of the window and associated with a set of operations;
• altering a state of the window based on the set of operations performed by each selected view;
• and
• obtaining a computing performance measurement based on a time required to alter the state of the window.

A non-transitory computer readable storage medium storing a computer program for benchmarking a computing performance of a window of an application running in a graphical user interface, the computer program comprising sets of instructions for:

• analyzing contents of the window to identify a plurality of views, each view describing a particular region of the window and associated with a set of operations;
• identifying, from the plurality of views, a set of related views for altering states of the window;
• altering a state of the window based on the set of operations performed by each identified related view;
• and
• obtaining a computing performance measurement based on a time required to alter the state of the window.