Group Bar The Task Bar Evolved Greg Smith
Group. Bar: The Task. Bar Evolved Greg Smith, Patrick Baudisch, George Robertson, Mary Czerwinski, Brian Meyers, Daniel Robbins, and Donna Andrews Microsoft Research
Key Problem: Task Switching Bannon et al. (1983) – information workers often switch between concurrent tasks n Rooms (Card & Henderson, 1987) n ¨ Working sets of windows Large displays & multimon lead to more open windows (4 > 12 > 16) n Interruptions lead to more task switching n OZCHI 2003: New Directions in Interaction Microsoft Research 2 of 22
Windows Task. Bar Problems n Task. Bar does not support task switching ¨ Many n operations required to make a switch Task. Bar does not scale well ¨ Grouping n by application rather than task Hypothesis: ¨ Movement, switching, & layout primitives at multi-window level can save time and effort OZCHI 2003: New Directions in Interaction Microsoft Research 3 of 22
Related Work n Virtual Desktop Managers ¨ Smalltalk Rooms Overview (1987) Project Views ¨ Rooms ¨ X-Windows ¨ Be. OS workspaces ¨ Linux – KDE desktops, etc. ¨ Win 32 ISV products: XDesk, Go. Screen, Flash Desktops, Desks. At. Will, etc. ¨ (not yet built into Windows) OZCHI 2003: New Directions in Interaction Microsoft Research 4 of 22
Related Work 3 D (Task Gallery) Time-Machine Computing OZCHI 2003: New Directions in Interaction Zooming (Pad++) Tiled (Elastic Windows) Microsoft Research 5 of 22
Group. Bar Design Points Familiar – build on Windows Task. Bar n Non-modal – not separate desktops n Lightweight UI – low-effort group creation and management n Leverage spatial memory: allow users to place tiles and groups for quicker recall n OZCHI 2003: New Directions in Interaction Microsoft Research 6 of 22
Demo OZCHI 2003: New Directions in Interaction Microsoft Research 7 of 22
Group. Bar Basics n Bar on any desktop edge n Resizable, auto-hide, always-on-top options n Multiple bars n One tile for each window OZCHI 2003: New Directions in Interaction Microsoft Research 8 of 22
Arranging Tiles n Drag tiles within bar to reorder n Drag tiles between bars for greater spatial separation OZCHI 2003: New Directions in Interaction Microsoft Research 9 of 22
Grouping Tiles (Main Theme) n Drag tile onto another tile to create group n Drag tile in/out to add/remove from group n Drag one of the last two tiles out to destroy group OZCHI 2003: New Directions in Interaction Microsoft Research 10 of 22
Dragging Subtleties n Move caret is straight n Insertion caret is curved toward group target n Target position decoupled from caret symbol to aid in target acquisition OZCHI 2003: New Directions in Interaction Microsoft Research 11 of 22
Group Appearance n Groups indicated by ¨ Subtle tile shape change ¨ Colored ¨ Green background frame group button OZCHI 2003: New Directions in Interaction Microsoft Research 12 of 22
Group Operations n Group button now offers a control surface n Click once to restore all n Click once to minimize all n Right-click for additional group operations OZCHI 2003: New Directions in Interaction Microsoft Research 13 of 22
Additional Group Operations Window Menu Group Menu OZCHI 2003: New Directions in Interaction Analogous to Window operations n Layout templates n ¨ Depend on display configuration ¨ Might depend on actual windows (not implemented) Microsoft Research 14 of 22
Overflow Strategies n Task. Bar ¨ Collapse n by app ¨ Multiple rows of tiles, buttons to ‘page’ Group. Bar ¨ Collapse by group ¨ Multiple bars Windows: paging buttons Group. Bar: collapsed group OZCHI 2003: New Directions in Interaction Microsoft Research 15 of 22
Longitudinal User Study n 5 participants n 7 -10 day study on their own work n Goal: initial understanding if users will use grouping for real work OZCHI 2003: New Directions in Interaction Microsoft Research 16 of 22
Results n Users did use grouping ¨ Average n 2. 5 groups of 2 windows Satisfaction ratings generally favorable ¨ Useful to drag to group ¨ Useful to close all windows in group at once ¨ Useful to remember layout ¨ Makes multiple monitors more useful OZCHI 2003: New Directions in Interaction Microsoft Research 17 of 22
Results (cont. ) n Negative satisfaction ratings for: ¨ More than one Group. Bar at a time ¨ Non-group windows minimize on group switch n 2 of 5 participants continued using Group. Bar after study OZCHI 2003: New Directions in Interaction Microsoft Research 18 of 22
Comparative User Study n Comparing Task. Bar and Group. Bar n 18 participants n 3 tasks consisting of 2 -3 documents each ¨ Planned n interruptions forced 5 task switches Triple monitor setup (3840 x 1024) OZCHI 2003: New Directions in Interaction Microsoft Research 19 of 22
Results n Borderline significant task time advantage ¨ 11. 7 n min vs 13. 25 min Satisfaction ratings significantly favor GB n GB unanimously preferred OZCHI 2003: New Directions in Interaction Microsoft Research 20 of 22
Future Work n Iterative design improvements ¨ Further studies for different display configurations and user tasks n Layout templates based on window use n Automatic grouping based on window use n Persistence OZCHI 2003: New Directions in Interaction Microsoft Research 21 of 22
Conclusions: Met Design Goals n Group. Bar provides basic task management ¨ Easy to group windows with drag and drop ¨ Single click task switching n Tasks shown with subtle extension to familiar Windows Task. Bar n Demonstrated ease of use, learnability, and user acceptance OZCHI 2003: New Directions in Interaction Microsoft Research 22 of 22
- Slides: 22