d.note Talk - CHI2010

Embed Size (px)

Citation preview

  • 8/9/2019 d.note Talk - CHI2010

    1/58

    Bjrn HartmannUniversity of California, Berkeley

    EECS, Computer Science Division

    Sean Follmer, Antonio Ricciardi,Timothy Cardenas, Scott Klemmer

    Stanford University HCI Group

    Revising User Interfaces ThroughChange Tracking, Annotations, and Alternatives

    d.note

  • 8/9/2019 d.note Talk - CHI2010

    2/58

    2

    Revision tools forword processing:

    Comments Change tracking Change

    visualization

  • 8/9/2019 d.note Talk - CHI2010

    3/58

    3

    Revision toolsfor user interfacedesigns?

    Revise:wrongmenu

    itemhighlighte

    d

    4differentmen

    u

    systems

    -confusing!

  • 8/9/2019 d.note Talk - CHI2010

    4/58

    Email survey of 10 UI interaction designers

    Ad hoc choices of digital, physical media: Printing out screens and sketching on them

    Posting digital screenshots to a wiki

    Using a bug tracking database and revision control

    4

  • 8/9/2019 d.note Talk - CHI2010

    5/58

    Designers prefer expressing changes to thecurrent design through sketching

    Designers prefer to capture history overtime through digital media

    5

  • 8/9/2019 d.note Talk - CHI2010

    6/58

    How can revision techniques ofcommenting, change tracking, and change

    visualization be applied to the revision ofuser interfaces?

    Are there techniques unique to UI design?

    How do interactive revision tools changedesigners practices?

    6

  • 8/9/2019 d.note Talk - CHI2010

    7/58

    Editable Graphical Histories, Kurlander,

    IEEE Visual Languages88Co-existence of paper and digital

    information for office work

    Outpost Design Histories,

    Klemmer, CHI02Capture & retrieval for web site design

    7

  • 8/9/2019 d.note Talk - CHI2010

    8/58

    Capturing Design HistoryEditable Graphical Histories, Kurlander,

    IEEE Visual Languages88Co-existence of paper and digital

    information for office work

    Outpost Design Histories,

    Klemmer, CHI02Capture & retrieval for web site design

    8

  • 8/9/2019 d.note Talk - CHI2010

    9/58

    DENIM, Lin, CHI00Sketch-based prototyping of web siteinformation architecture

    Topiary, Li, UIST04Prototyping of mobile, location-awareapplications

    9

  • 8/9/2019 d.note Talk - CHI2010

    10/58

    DENIM, Lin, CHI00Sketch-based prototyping of web siteinformation architecture

    Topiary, Li, UIST04Prototyping of mobile, location-awareapplications

    10

  • 8/9/2019 d.note Talk - CHI2010

    11/58

    Annotation ToolsPaper Augmented Digital Documents

    Guimbretiere, UIST03Digital pen marks on paper change

    underlying document

    ModelCraft, Song, UIST06Digital pen marks on 3D objects changeCAD model

    11

  • 8/9/2019 d.note Talk - CHI2010

    12/58

    Annotation ToolsPaper Augmented Digital Documents

    Guimbretiere, UIST03Digital pen marks on paper change

    underlying document

    ModelCraft, Song, UIST06Digital pen marks on 3D objects changeCAD model

    12

  • 8/9/2019 d.note Talk - CHI2010

    13/58

    Interaction design specifies bothappearance and behavior.

    What tools do exist for revising each?

    13

  • 8/9/2019 d.note Talk - CHI2010

    14/58

    14

    Microsoft & Pixar, from Buxton, Sketching User Experiences.

  • 8/9/2019 d.note Talk - CHI2010

    15/58

    Diff tools show two versions side-by-side

    15

  • 8/9/2019 d.note Talk - CHI2010

    16/58

    Goals:

    Enable rapid expression of changes,at different levels of representation

    Communicate clearly what changed (and how)

    16

  • 8/9/2019 d.note Talk - CHI2010

    17/58

    Goals:

    Enable rapid expression of changes,at different levels of representation

    Communicate clearly what changed (and how)

    Approach: Use sketching as input

    Develop visual revision language

    Revision of both design and source

    17

  • 8/9/2019 d.note Talk - CHI2010

    18/58

    18

    d.note: revision techniques for UI prototypes

  • 8/9/2019 d.note Talk - CHI2010

    19/58

    19

    d.note: revision techniques for UI prototypes

  • 8/9/2019 d.note Talk - CHI2010

    20/58

    20

  • 8/9/2019 d.note Talk - CHI2010

    21/58

    21

  • 8/9/2019 d.note Talk - CHI2010

    22/58

    22

    Adding States Adding Transitions Adding Graphics

    Removing States Removing Transitions Adding Comments

  • 8/9/2019 d.note Talk - CHI2010

    23/58

    23

    d.note: revision techniques for UI prototypes

  • 8/9/2019 d.note Talk - CHI2010

    24/58

    24

  • 8/9/2019 d.note Talk - CHI2010

    25/58

    25

  • 8/9/2019 d.note Talk - CHI2010

    26/58

    26

  • 8/9/2019 d.note Talk - CHI2010

    27/58

    27

    Alternatives share:

    Incoming transitions

    Alternatives define:

    Screen graphics Outgoing transitions

  • 8/9/2019 d.note Talk - CHI2010

    28/58

    28

    Gestures are recognized by Wobbrocks $1recognizer with multiple templates for eachgesture.

  • 8/9/2019 d.note Talk - CHI2010

    29/58

    29

    How does expressing revisions interactively

    compare with sketching on static images?

  • 8/9/2019 d.note Talk - CHI2010

    30/58

    30

    Within-subjects design, n=12

    Each participant revises two prototypes:

    one with d.note,one by sketching on a static image

    Static condition assigned first,

    prototype counterbalanced

  • 8/9/2019 d.note Talk - CHI2010

    31/58

    31

  • 8/9/2019 d.note Talk - CHI2010

    32/58

    32

  • 8/9/2019 d.note Talk - CHI2010

    33/58

    Digital Camera Navigation Prototype

    33

  • 8/9/2019 d.note Talk - CHI2010

    34/58

    Keychain Photo Viewer Prototype

    34

  • 8/9/2019 d.note Talk - CHI2010

    35/58

    35

  • 8/9/2019 d.note Talk - CHI2010

    36/58

    36

  • 8/9/2019 d.note Talk - CHI2010

    37/58

    37

  • 8/9/2019 d.note Talk - CHI2010

    38/58

    38

  • 8/9/2019 d.note Talk - CHI2010

    39/58

    39

  • 8/9/2019 d.note Talk - CHI2010

    40/58

    With d.note, participants: wrote less

    (2.3 vs 9.8 mean text comments, p

  • 8/9/2019 d.note Talk - CHI2010

    41/58

    Ability to test proposed changes(7 of 12 participants)

    Ability to make functional changes (6)

    Less cluttered than sketching alone (2)

    Notation easier to interpret (1)

    41

  • 8/9/2019 d.note Talk - CHI2010

    42/58

    Commenting is more difficult(3 participants)

    Tool has steeper learning curve (2)Danger of getting stuck on details (2)

    42

  • 8/9/2019 d.note Talk - CHI2010

    43/58

    It was easier to write/sketch withproduction software than with our tool.

    Formal notation reduced the need forexplanations.

    43

  • 8/9/2019 d.note Talk - CHI2010

    44/58

    Immediate testing enabled discovery oferrors and ambiguities.

    44

  • 8/9/2019 d.note Talk - CHI2010

    45/58

    Participants focused on single-staterevisions because the tool gave them moreleverage to do so.

    45

  • 8/9/2019 d.note Talk - CHI2010

    46/58

    46

    How well can others interpret

    interactively created revisions?

  • 8/9/2019 d.note Talk - CHI2010

    47/58

    47

    New set of participants (n=8) rates theoutput of the last study

    Each participant rates 4 diagrams:camera revised with d.notecamera revised without d.note

    keychain revised with d.note

    keychain revised without d.note

  • 8/9/2019 d.note Talk - CHI2010

    48/58

    48

  • 8/9/2019 d.note Talk - CHI2010

    49/58

    49

    no significant difference - t(30)=0.25, p=0.15

    error bars show standard error

    meannumberofrevisionslisted

    interactive revision

  • 8/9/2019 d.note Talk - CHI2010

    50/58

    50

    significant difference - t(30)=1.90, p

  • 8/9/2019 d.note Talk - CHI2010

    51/58

    51

    Changes are concrete and specific(4 of 8 participants)

    Revisions contain proposed solutions (3)

  • 8/9/2019 d.note Talk - CHI2010

    52/58

    52

    Visual clutter in regions of dense changes(5 participants)

    Hard to glean motivation for changes (4)

  • 8/9/2019 d.note Talk - CHI2010

    53/58

    53

    Reduce visual complexity inrevised diagrams

    Encourage capture of rationale

  • 8/9/2019 d.note Talk - CHI2010

    54/58

    54

  • 8/9/2019 d.note Talk - CHI2010

    55/58

    An operationalization of common revisionoperations for user interfaces.

    55

  • 8/9/2019 d.note Talk - CHI2010

    56/58

    An operationalization of common revisionoperations for user interfaces.

    New revision operationsunique to user interfaces:testing proposed changes immediately

    specifying alternatives to existing functionality

    56

  • 8/9/2019 d.note Talk - CHI2010

    57/58

    57

    Interactive revision offer benefits:immediate testing and formal notation lead toincreased clarity of revisions

    For a tradeoff:formal notation discouraged broad, conceptual changes;capturing motivation for changes

    Future work:encourage revisions at multiple levels of abstraction

  • 8/9/2019 d.note Talk - CHI2010

    58/58

    [email protected]

    http://hci.berkeley.edu/mailto:[email protected]://hci.berkeley.edu/mailto:[email protected]:[email protected]