Our CHI 2007 paper  describes a controlled experiment comparing four commonly-used copy-and-paste techniques (keyboard shortcuts, a context menu, drag-and-drop and a specific X Window technique) under four window management conditions: non-overlapping windows, partially overlapping windows, and partially overlapping ones with one of two new window-management techniques: restack (aka rock) and roll.
Results from this experiment show that these two window-management techniques significantly reduce the task completion time for all copy-and-paste techniques. They also show that X Window copy-and-paste is faster than the three other techniques under the four window management conditions.
The following videos illustrate this work by showing different ways of copying text between partially overlapped windows:
|The naive way||The users clicks on a partially covered Web browser, which raises it. He selects some text. He then clicks on the editor window, which raises it, and pastes the text using the middle mouse button.|
|A better way||The users starts a text selection in a partially covered Web brower, which raises this window. He then clicks on the editor window to raise it and pastes the selection using the middle mouse button.|
|An even better way||The users starts a text selection in a partially covered Web brower, which raises this window. He then pastes the selection in the partially covered editor using the middle mouse button, which raises that window.|
|Rock!||When the selection starts, the Web browser is automatically raised. When it stops, the browser is automatically put back to its previous place in the stacking order.|
|Roll!||When the selection starts, windows covering the target window are automatically rolled to uncover it. When the selection ends, these windows roll back to their original condition.|
The restack and roll techniques are available as part of the Metisse window system. Source code for the experiment described in  is also available.
This work has been partially supported by the French ACI Masses de données (Micromégas project, 2003-2006).