Manipulation of visual information refers to the activities performed to restructure information organization on the screen and to update screen contents.
In current systems, tasks that require interaction with information contained in multiple windows are difficult because operations are performed independently, one window at a time. This often makes it tedious to manage multiple windows. According to an early study by Bury et al. , multiple windows improve user performance only if users do not spend a lot of time for window housekeeping (i.e. opening, moving, resizing, etc.). This study indicates that window housekeeping activities need to be decreased in order to reduce users' cognitive overhead. According to my observations, the window layout is typically kept unchanged. Very few move and resize operations were performed by participants even though their tasks changed. This result also indicated that making layout rearrangements is quite tedious.
People typically organize the papers on their desk into piles, and move all of them simultaneously. Malone  finds that users like to group items spatially. Application of window operations on groups of windows is likely to improve users' information manipulation performance dramatically. Composition of multiple windows into groups as separate entities, and application of window operations on these group windows facilitate scalability in terms of the number of operations.
Current windowing systems fail to support users in performing their tasks. As Funke et al.  suggest, windowing systems should support users to integrate, organize, compare, distill, summarize and apply the information. Besides, users also need to restructure information organization on their screen in order to meet their new information demands as required by their evolving tasks. This is quite tedious in current systems due to the independence of window operations.