ZOIL - A Cross-Platform User Interface Paradigm for

Presented at: Personal Information Management 2008 (PIM 2008), CHI 2008 Workshop, April 5-6, 2008, Florence, Italy
ZOIL – A Cross-Platform User Interface Paradigm for
Personal Information Management
Hans-Christian Jetter
Werner A. König
Jens Gerken
Harald Reiterer
Human-Computer Interaction Group, University of Konstanz
Universitätsstraße 10, 78457 Konstanz, Germany
{jetter,koenig,gerken,reiterer}@inf.uni-konstanz.de
ABSTRACT
[35], but also “web-based”, “nomadic” [30], “ubiquitous”
[5] or “social” [14]. Most functionality has been deployed
to the user as local “desktop applications”, but some also as
“objects”, “templates” and “views” [27], as “activities” [5],
or as “web widgets”, “web services” or “web applications”
based on dynamically generated hypertext [1]. Even the
local file system has changed its face from a simple
hierarchical storage structure into a versatile database with
indices of content and metadata [11, 12] to allow new ways
of querying and accessing personal information.
In this paper we introduce the novel user interface paradigm
ZOIL (Zoomable Object-Oriented Information Landscape).
ZOIL is aimed at unifying all types of local and remote
information items with their connected functionality and
with their mutual relations in a single visual workspace as a
replacement of today’s desktop metaphor. This workspace
can serve as an integrated work environment for traditional
personal information management (PIM), but can also be
used for PIM tasks in a wider sense. By formulating ZOIL’s
fundamental design principles we describe the interaction
style, visualization techniques and interface physics of a
ZOIL user interface. Furthermore we discuss ZOIL’s ability
to provide nomadic PIM environments for mobile and
stationary use.
Over the decades this great diversity of usage scenarios and
design goals has left many traces in today’s design of our
personal information management tools. Although the
desktop metaphor is still thought to be the centerpiece of
PIM activities and work environments, our PCs, laptops,
smart phones or PDAs are packed or even bloated with a
multitude of non-interoperable specialized PIM applications
and websites which carry out the actual work (e.g. Google
Calendar, Facebook, BSCW or Microsoft Outlook). Most of
these use incompatible storage formats and inconsistent
interaction models (e.g. desktop GUI applications vs.
hypertext-driven web applications [29]) which have further
hollowed out the role of the desktop metaphor [20].
Author Keywords
personal information management, post-WIMP
interfaces, information visualization, zoomable
interfaces, object-oriented user interfaces.
user
user
ACM Classification Keywords
H5.2. Information interfaces and presentation (e.g., HCI):
User Interfaces.
For many PIM users these inconsistencies have led to an
almost paralyzing amount of necessary workarounds and to
a destructive degree of complexity and “information
fragmentation” [24]. Content and functionality are scattered
over dozens of applications, websites, storage formats,
interaction models or devices with each one posing an
individual challenge to the user’s cognitive skills. Jef
Raskin has identified this critical weakness of today’s
“mazelike” interface (Raskin, [32]) as one of the main
problems standing between current technology and
tomorrow’s “humane interface”. He regards “fundamental
changes in the design of human-machine interfaces” as
inevitable, since “nothing less will do”. The prospects of
such a fundamental change have led researchers to suggest
designs for the “integrated digital work environments” of
tomorrow which go “beyond the desktop metaphor” and
open all new perspectives for personal information
management [23]. The goal is to design a general-purpose
interface suitable for many different devices which unifies
all kinds of content and functionality under a consistent
interaction model while leaving the user the possibilities to
INTRODUCTION
Throughout its history Human-Computer Interaction (HCI)
and related disciplines have been striving to research and
design novel usable user interfaces that unify qualities like
learnability, suitability for the task, “intuitiveness” or
attractiveness. Thereby the proposed designs have always
been influenced by the then-available technology and the
personal, organizational or societal values of the time. Thus
the user interface (UI) of the personal computer has
undergone many changes to become “graphical”,
“intuitive”, “object-oriented” [10] or “direct-manipulative”
1
Konstanzer Online-Publikations-System (KOPS)
URN: http://nbn-resolving.de/urn:nbn:de:bsz:352-opus-75367
URL: http://kops.ub.uni-konstanz.de/volltexte/2009/7536
establish own workflows, data structures or views on her
information space.
concept of ZOIL portals which form the backbone of every
ZOIL-based UI (see following section).
At the Human-Computer Interaction Group (HCIG) of the
University of Konstanz we are designing and researching
such a novel visual user interface concept named “ZOIL”
(Zoomable Object-Oriented Information Landscape). As an
application- and platform-independent UI concept ZOIL is
aimed at unifying all types of local and remote information
items with their connected functionality and with their
mutual relations in a single visual workspace under a
consistent interaction model. This visual workspace is
named the “information landscape” and can serve as an
integrated work environment for traditional PIM tasks, but
can also be used for PIM tasks in a wider sense like
navigating public information spaces on the Web (e.g.
digital libraries, social networking websites or georeferential content in maps).
Jef Raskin’s vision of the “ZoomWorld” and the “Zooming
Interface Paradigm” from 2000 has been a further important
source of inspiration: “The zooming interface paradigm can
replace the browser, the desktop metaphor, and the
traditional operating system. Applications per se
disappear.” (Raskin, [32]). Thereby Raskin’s concepts of
“unification” and “commands” have similarities to objectoriented user interfaces which have been discussed by Theo
Mandel in 1994 [27] and Dave Collins in 1995 [10].
At the current stage of our project we have formulated five
design principles for ZOIL user interfaces and have applied
them in mock-ups and Java prototypes for document
management [26] (see figure 3) and digital libraries [20]
(see figure 6). The design principles are based on our
previous research work on Zoomable User Interfaces
(ZUIs) [19, 20, 25], visual information seeking [17] and
information visualization (IV) [33]. Furthermore they have
drawn from a broad scope of existing designs and theories
in the HCI, IV and PIM literature.
By describing these five design principles we will introduce
the ZOIL paradigm as subject of our ongoing research work
to the scientific community. Furthermore we will outline
the potential of ZOIL user interfaces to replace the desktop
metaphor and traditional application-oriented PIM environments. The role of ZOIL user interfaces in reducing
information fragmentation and providing usable nomadic
PIM solutions will be a further topic of our discussion,
which will then be concluded by outlining our next steps
and future work.
PREVIOUS AND RELATED WORK
The first use of a zoomable virtual canvas to visually access
and spatially arrange information items dates back to
William Donelson’s and Richard A. Bolt’s seminal “Spatial
Data Management System (SDMS)” from 1978 [13]. “Pad”
of Perlin & Fox and “Pad++” of Bederson & Hollan
provided pioneering theories and more elaborated designs
of such ZUIs in 1993 and 1994: “Pad” was based on the
assumption that navigation in information spaces is best
supported by tapping into our natural spatial and
geographical ways of thinking [6, 31]. Data objects are
organized in space and scale and users can navigate them
by performing zooming and panning operations. “Pad” has
also introduced the concepts of semantic zooming and
portals. We have merged Pad’s “portals” and “portal filters”
with the “see-through interface” of Bier et al. [9] and the
“dynamic queries” of Ahlberg et al. [3] to create the
“Unification” is also discussed by David R. Karger and
William Jones in the context of PIM and the “Haystack”
work environment [23, 24]. Haystack uses a RDF-based
semi-structured data model for organizing personal
information items. A related approach can also be found in
the DeepaMehta semantic desktop project [36].
Concerning the implementational aspects of ZUIs, Ben
Bederson’s Jazz and Piccolo frameworks provide valuable
input on architectures and data models [8]. A very
impressive implementation of a ZUI based on Microsoft
Live Labs’ “Seadragon” was presented by Blaise Aguera y
Arcas at the TED conference 2007 [2]. This demonstration
has influenced our decision to develop a web-deployed
ZOIL UI software framework based on Microsoft .NET and
WPF as part of our research activities.
As will be discussed in more detail in the following section,
the ZOIL design principles furthermore reflect some of the
findings of William Jones and Jamie Teevan about the
nature of PIM, personal spaces of information (PSI) and
personal information collections (PIC) [22]. These essential
PIM concepts and the growing body of knowledge about
PIM practice will be guiding our future design and
evaluation of the ZOIL paradigm (see section “Conclusion
& Future Work”).
THE ZOIL USER INTERFACE PARADIGM
To describe the interaction style, visualization techniques
and interface physics of a ZOIL user interface we have
formulated five ZOIL design principles which will be
introduced in the following sections.
ZOIL Design Principle 1: Object-Oriented User Interface
In [22, p.7] Jones and Teevan introduce the “information
item” as a fundamental concept for the consideration of
PIM. An “information item” (e.g. a real world printed
document or handwritten note, an email message, a web
page or a reference to a web page) is a “packaging of
information in a persistent form […]” and with an
associated “information form” or “information type” (i.e.
items have types like “paper document”, “electronic
message”, “web page” or “web bookmark"). This
information form is “determined by the constellation of
tools and applications that make it possible to manipulate
the item.”
system in the early 1990s. At that time OS/2’s objectoriented approach was perceived as the direct competitor to
the application-orientation in the GUI of Microsoft
Windows and was regarded as more “intuitive” and
“natural” by many HCI researchers [27]. However, due to
the rapidly growing popularity of the Windows operating
system and the advent of the hyperlink-driven World Wide
Web, OOUIs have never gained much attention and
momentum in UI design or PIM since then.
A more formal model of this abstraction can be easily
achieved by following an object-oriented (OO) approach:
All information is contained in numerous interrelated
information items (or objects) of different information
forms (or classes). Each information form can be formally
specified by a class definition and each information item
can be considered as an instance of a class following this
class definition. The attributes section of a class definition
defines a template for the possible content, relations and
metadata of the individual information items and the
methods section of the class definition specifies all possible
ways of interacting with the item or manipulating it.
This object-oriented model comes in handy when
considering Jones and Teevan’s observation that although
“a person’s interactions with an information item vary
greatly depending upon its form” “there are many essential
similarities in the way people interact with information
items, regardless of their form.” By using the OO
mechanism of inheritance it is possible to model such
essential similarities between different information forms in
a common base class. If properly applied this leads to a
class hierarchy of information forms which integrates very
different types of information while preserving a maximum
degree of consistency in attributes and methods. At first
glance this might only appear interesting for programmers.
However, such a consistent class hierarchy can also have a
strong impact on usability and UI design when following
the approach of “Object-oriented User Interfaces”.
Application-Oriented
User Interfaces
Object-Oriented User
Interfaces
Application consists of an
icon, primary window,
and secondary windows
Product consists of a
collection of cooperating
objects and views of objects
Icons represent running
applications
Icons represent objects that
may be directly
manipulated
Users must start
application before
working with objects
Users open objects into
views
Provide users with
function needed to
perform a task
Provide users with supplies
needed to perform a task
Rigid structure-by
function
Flexible structure-by object
Users must follow the
application structure
User may perform in their
own way or innovate
Many applications
required – one per task
Few objects – more reuse of
the same object in many
tasks
Figure 1. Domain model and data sources of a ZOIL-based UI.
One of the key concepts of OOUIs is to deploy functionality as a collection of cooperating objects and views of
objects which are directly integrated as new components
into a common visual workplace. Instead of creating
monolithic applications running in isolated windows with a
specialized interaction model, rigid workflows and noninteroperable data formats, new functionality is seamlessly
integrated into the existing desktop environment. This
preserves a high degree of consistency and interoperability
on the UI (see table 1 from Theo Mandel [27]). Unlike in
monolithic applications, tasks in OOUIs are then carried out
in the desktop environment by direct manipulations in a
very flexible “model world” interface instead of indirect
interactions in specific applications with rigidly structured
“conversational interfaces” and dialogs [18]. Therefore
OOUIs appear especially suitable for such complex and
unstructured “knowledge work” [10] like PIM where
workflows and tasks are highly user-specific and cannot be
easily automated or supported by conventional dialogs.
Only fragments of the original OOUI principles have found
their way into today’s desktop metaphor and applications,
e.g. the object-action sequence of context/pop-up menus or
the “pick, drag and drop” interaction with icons. ZOIL
takes up OOUI concepts in their original sense to provide
the user with a flexible though consistent direct
manipulation interface for PIM.
Table 1. Application-oriented UI vs. OOUI (taken from [27]).
Object-oriented user interfaces (OOUIs) were introduced to
the PC GUI in connection with IBM’s OS/2 operating
3
For this reason all local and remote information items in
ZOIL are regarded as objects of different classes as
described in the aforementioned OO model of information
in PIM. The class of an object defines its attributes and
metadata fields (e.g. “From:”-field of a mail or GPS
coordinates of a photo) and the possible relations to other
items (e.g. a mail is sent by a person, a mail is sent to a
person, a person can be a friend of other persons).
Furthermore the class defines the available user interaction
possibilities (or methods) connected with the object and
their visibility and presentation on the user interface
depending on the available screen estate and the currently
active views and filters (see figure 1). The class definitions
in ZOIL thereby resemble a basic domain model of PIM
that constitutes a “model world” interface defining visible
attributes, relations and functionality of information items
including their presentation and behavior on the user
interface. OO inheritance and polymorphism are then used
to achieve a maximum degree of consistency in the classes
and therefore also in the ways the user can interact with the
system and manage different information forms.
However, an OOUI with a basic domain model might still
be too restrictive to satisfy all user needs in PIM. As David
Karger points out in [25] developers can hardly predict
which information objects need to be stored, retrieved, and
viewed or what relationships and attributes are worth
storing and recording in day-to-day practice. Thus the
domain model within a ZOIL user interface should be
extensible and customizable by the user through easily
adding new information forms with new functionality or
adding further properties, relations and views anytime
during runtime.
ZOIL Design Principle 2: Semantic Zooming
In a ZOIL user interface all information items from all data
sources appear as visual representations at one or more
places in the information landscape. However, unlike in
today’s desktop metaphor, this landscape is not limited to
the visible screen size but resembles a virtual canvas of
infinite size and resolution (as known from [31] and [32]).
All items and their connected functionality can be accessed
by panning to the right spot in the information landscape
and zooming in. ZOIL thereby uses “semantic zooming”
[31] which means that the geometric growth in display
space is not only used to render more details, but to reveal
that content and functionality which is of most use to the
user. If the user zooms into more complex items like
documents, drawings or spreadsheets they become editable
on-the-spot without the need to open a dedicated
application window (see figure 2). Thus the available
functionality is always coupled with the information item
itself as it is proposed by object-oriented user interface
design.
Zooming out of the information landscape leads to a
decrease of display space. The visual representation of an
information item gradually collapses to an icon or glyph.
Eventually an information item is represented as a single
remaining pixel. Together with the neighboring pixels this
pixel can then be perceived as a cluster or a part of an
overview visualization aggregating a large amount of items.
Figure 2. Semantic zooming into an information item reveals
not only more content but also the necessary functionality.
An important benefit of using semantic zooming to navigate
the information space is the natural “feel” and the
“intuitiveness” of zooming because of its similarity to the
principles of the physical world. However, a pure ZUI has
some disadvantages because of the absence of peripheral
views. This will be part of our future research work and is
discussed in the last section of this paper.
ZOIL Design
Visualization
Principle
3:
Nested
Information
Since the number of information items in PIM can easily
add up to thousands or millions of items, it is necessary to
find ways to structure large amounts of items and to
efficiently access them according to the user’s current
information need and activity. In current practice we
typically use physical piles of papers on our desktops,
electronic folders and subfolders on our hard disk or the
inbox of our email client to establish such “islands of
relative structure and coherence.” [22, p.12] Jones and
Teevan refer to these islands as “personal information
collections” (PICs) where “people have made some
conscious effort to control both the information that goes in
and also, usually but not necessarily, how this information
is organized”. Thereby PICs should be considered “not as
defined by technical format or application, but rather as
defined by activities of people […].”
In ZOIL, equivalents to such PICs can be created
persistently as “portals” at arbitrary locations in the
information landscape by specifying a size and location in
the landscape and assigning the desired information items
and a visualization type (e.g. geographic maps, time lines,
networks, tree views or simply an empty canvas) to the new
portal. These visualizations are then used to display each
contained information item according to the items’
properties or relations. For example a collection of photos
could appear as zoomable thumbnails on a geographic map
using GPS coordinates or in temporal order within a
timeline using their timestamp. Since ZOIL user interfaces
make use of an object-oriented domain model, different
kinds of information items can be freely combined within a
single portal using their common base classes. This enables
the user to place and manage different forms of information
in a single PIC without being limited to certain data formats
a rectangular region of a map visualization is selected and is
transformed into a portal displaying all documents within
the selected region in a scatter plot.
or application boundaries as is often the case in today’s
application-oriented work environments.
To control the information that goes into such portals a
ZOIL UI offers different direct manipulation techniques.
For example the user can manually assign items or item
collections to a portal by using a pick, drag and drop
interaction as an equivalent to copying in today’s desktop
metaphor. Furthermore dynamic queries [4] can be used to
formulate search queries on the information space (e.g.
desktop and web search) that automatically fill the portal
with the result set of information items. Figure 6 illustrates
how metadata like “Year” and “Persons” could be used to
search for DVDs in the information space and to iteratively
specify the desired items. In this example a range slider is
used to specify the desired years of publication.
Furthermore an incremental keyword search is used to find
all DVDs with a certain cast member by typing the first
letters of the person’s name.
Using the information landscape to create and manage
multiple portals enables the user to create persistent views
on arbitrary subsets of the information space dedicated to
the user’s typical information needs or current activities.
For example the user could create a portal that is
prominently positioned on the information landscape and
that contains a zoomable table view of recent emails similar
to today’s inboxes in email clients. Figure 3 shows a
screenshot of a ZOIL-based work environment for
document management which contains six different portals
to access documents: A person-oriented access is provided
by the floor plan below “Persons” which allows exploring
the documents of an author according to the location of the
author’s office. The calendar visualization on the top of the
screen positions the documents according to their creation
and modification date and allows a time-oriented zoom into
their content or to keep track of recently received
documents. Furthermore the user can use the world map
under “Location” to zoom into scientific papers according
to the geographic location of a conference (see figure 4).
Further visualizations ranging from hierarchical treemaps
[7] to social networks [14, 33] or zoomable tables [19] can
be integrated into the information landscape by
downloading them as modular “plug-ins” as it is known
from today’s web browsers. Thereby it is important to
notice that information items in ZOIL are not located at a
A further way to formulate complex queries is the approach
of “nested information visualization”. This approach
enables the user to select arbitrary rectangular regions in the
information landscape or in portals (e.g. with a bounding
box, see figure 4) to create a new portal within the existing
portal that only contains the items from the selected region.
As suggested by Perlin & Fox in [31] portals can also be
used like the “magic lenses” of Bier et al. [9] to temporarily
or persistently view the content of a portal with different
visualizations and filter criteria. Figures 4 and 5 show how
Figure 3. A ZOIL-based work environment for document management
(The dotted boxes illustrate the semantic zooming in the information landscape).
5
single spot in the information landscape. Unless they are
filtered out with dynamic queries and nesting or unless they
are manually assigned exclusively to one portal, each item
in the information space is placed in each of the portals.
This allows a flexible complementary use of different
visualizations depending on the current information need.
constantly changing, an information system can neither
provide an complete overview of a person’s PSI nor can it
process the whole PSI to search or manage all PICs or
information items within. However, for facilitating PIM
tasks it appears reasonable that a digital PIM environment
should try to integrate and provide as much of the PSI and
the contained PICs as possible – regardless of the
information form or source.
Figure 4. A zoom into a map visualization allows a
geographical access to documents. Furthermore a bounding
box is drawn to select the Eastern regions of the US.
Figure 5. The bounding box in figure 4 is transformed into a
portal visualizing the contained documents in a scatter plot.
ZOIL Design Principle 4: The Information Space as
Information Landscape
All PIM tasks of a user take place within the user’s single
“personal space of information” (PSI) that “includes all
information items that are, at least nominally, under that
person’s control.” [22, p. 10] A user’s PSI spans the virtual
and physical world and contains a great variety of
information forms ranging from the physical book in the
bookshelf to the text message on the mobile phone. Since
the PSI’s content and boundaries are uncertain and
Figure 6. A mock-up of a ZOIL-based library catalog. The
amount of displayed information is reduced by specifying the
desired years of publication (“Jahr”) and the name of a
desired cast member (“Personen”).
interface and interaction style regardless of the employed
device. ZOIL therefore leads to a significant improvement
of the cross-platform consistency leading to high a degree
of “horizontal usability” [34].
The information landscape in ZOIL can be regarded as such
an attempt to visually map as much as possible of the user’s
PSI and to provide a visual information workspace unifying
heterogeneous information items of different forms and
from different data sources and devices (e.g. the file system,
remote servers, web services, currently connected devices).
Thereby the default layout of visualizations and content in
the information landscape is predefined by the UI designer.
However, the concept of portals and direct manipulation
enables the user to customize or create own layouts
according to the user’s typical information needs or her
personal view of the PSI. As described in the previous
section, important views or visual queries on the
information space can be stored at prominent places for
later use or subsets of information items can be dragged to
arbitrary locations to arrange or annotate them freely on a
canvas. This allows the user to design the appearance of the
work environment, for example to create regions in the
information landscape that can be used as visual equivalents
to to-do lists or bookmarks supporting typical tasks in PIM.
Furthermore sketching, drawing or annotation tools can be
used to create own visual arrangements and visual
landmarks to facilitate information access and spatial
orientation or to allow zoomable presentations of content
within the information landscape. Thus the information
landscape can also be used as a versatile direct
manipulation authoring and annotation tool comparable to
the multiscale editor “MuSE” of Furnas & Zhang [15]. In
combination with pen-based devices like Tablet PCs or in
environments with large displays this allows the use of a
ZOIL-based user interface not only as a PIM tool, but also
as a novel artistic or collaborative work environment.
Figure 7. ZOIL as a cross-platform user interface paradigm
deploying the personal information space to various devices.
Figure 8. ZOIL-based user interfaces on a PDA, Tablet PC,
Multitouch Display and a large high-resolution display.
ZOIL Design Principle 5: Nomadic Cross-platform User
Interfaces
In our vision of nomadic personal information management
(see figure 7) the user can easily access her/his personal
information from any device connected to the World Wide
Web. Similar to current web-based email accounts the user
simply logs into a web server which then provides the
user’s personal information in a familiar ZOIL-based rich
Due to the nature of ZUIs a ZOIL user interface scales to
different display sizes and screen resolutions. This
scalability facilitates the preservation of a consistent user
7
internet application. This creates the illusion of a pervasive
nomadic PIM environment which is not bound to specific
operating systems or devices but is traveling with the user
from device to device.
In course of our “permaedia” project [21] we will research
how ZOIL-based user
interfaces
support this
complementary use of such different devices like small
display devices (e.g. Ultra Mobile PCs, PDAs or smart
phones), desktop and living room devices (e.g. Tablet PCs,
Home Theater PCs, set-top boxes) and large high resolution
displays (see figure 8).
CONCLUSION & FUTURE WORK
With the ZOIL user interface paradigm we have outlined a
novel approach to the design of integrated work
environments for PIM. By formulating the five fundamental
ZOIL design principles and illustrating them in mock-ups
and prototypes we have made a first step towards an
alternative to the traditional application-oriented desktop
metaphor. We believe that ZOIL can evolve into the
foundation of a new generation of cross-platform PIM user
interfaces with high usability and an intriguing user
experience. In particular the web-based deployment of
ZOIL user interfaces as rich internet applications would
open new possibilities for a pervasive and nomadic personal
information management on a wide range of devices under
a consistent interaction model.
For this reason we will develop provisional user interface
libraries, input device managers and backend interfaces
based on C# and .NET/WPF to allow a quick realization
and evaluation of elaborated ZOIL-based prototypes for
different stationary and mobile devices. The results from
user tests will be used to optimize our designs in various
contexts of use and to refine the formulation of the
fundamental ZOIL design principles. In particular
following important research questions remain unanswered
at this stage and will be part of our future work:
How does the visual-spatial orientation in a ZUI stand the
test of day-to-day use? The user’s orientation and
navigation in ZUIs will be subject to empirical research and
close examination as this ability is critical for the benefits
of a ZUI and has not been researched sufficiently yet [16].
How can the ZOIL paradigm be enhanced by integrating
peripheral views or polyfocal navigation? A user interface
that is purely based on zooming and panning in the
information landscape lacks permanently visible indicators
for events like “an email has arrived” or lacks free floating
regions of the screen for activities like instant messaging or
video chats. Furthermore comparing items or transferring
parts of one item into another should be facilitated by
offering multiple simultaneous views on different locations
in the landscape (e.g. by splitting the screen, by collapsing
and expanding regions of the landscape or by using
polyfocal distortion).
How can the ZOIL paradigm be enhanced through
alerting? As Jones and Teevan point out in [22, p. 9]
deferring processing until later is a frequent activity in PIM.
To this day ZOIL does not contain possibilities to easily
mark an information item for later processing and to keep
track of these items, their changes or deadlines. Alerting the
user about new arrivals, changes or at prescribed points in
time should be integrated without impairing the consistency
and visual character of the ZOIL UI.
Although the ZOIL paradigm is still at a very early stage,
we are convinced of its potential for the future. In [28]
Moran & Zhai have postulated seven dimensions in which
future work environments should emancipate themselves
from the traditional desktop metaphor. Regarding the new
possibilities that the ZOIL paradigm introduces to user
interface design we are convinced that ZOIL will be able to
contribute innovative solutions in all these dimensions.
REFERENCES
1. Adler, S. WebOS: say goodbye to desktop applications.
netWorker 9, 2005, 18-26.
2. Talks: Blaise Aguera y Arcas Photosynth demo. http://www.ted.com/.
Jaw-dropping
3. Ahlberg, C., Williamson, C. and Shneiderman, B.
Dynamic queries for information exploration: an
implementation and evaluation. In CHI '92: Proceedings
of the SIGCHI conference on Human factors in
computing systems, ACM Press (1992), 619-626.
4. Ahlberg, C. and Shneiderman, B. Visual information
seeking using the FilmFinder. In CHI '94: Conference
companion on Human factors in computing systems,
ACM Press (1994), 433-434.
5. Bardram, J.E. From Desktop Task Management to
Ubiquitous Activity-Based Computing. In Beyond the
desktop metaphor : designing integrated digital work
environments, MIT Press (2007), Cambridge, Mass.,
223-259.
6. Bederson, B.B. and Hollan, J.D. Pad++: a zooming
graphical interface for exploring alternate interface
physics. In UIST '94: Proceedings of the 7th annual
ACM symposium on User interface software and
technology, ACM Press (1994), 17-26.
7. Bederson, B.B., Shneiderman, B. and Wattenberg, M.
Ordered and quantum treemaps: Making effective use of
2D space to display hierarchies. ACM Trans. Graph 21,
ACM Press (2002), 833-854.
8. Bederson, B.B., Grosjean, J. and Meyer, J. Toolkit
Design for Interactive Structured Graphics. IEEE Trans.
Softw. Eng 30, IEEE (2004), 535-546.
9. Bier, E.A., Stone, M.C., Pier, K., Buxton, W. and
Derose, T.D. Toolglass and magic lenses: the seethrough interface. In SIGGRAPH '93: Proceedings of
the 20th annual conference on Computer graphics and
interactive techniques, ACM Press (1993), 73-80.
24. Karger, D.R. and Jones, W. Data unification in personal
information management. Communications of the ACM
49, ACM Press (2006), 77-82.
10. Collins, D. Designing object-oriented user interfaces.
Benjamin Cummings, Redwood City, CA, USA, 1995.
11. Cutrell, E., Dumais, S.T. and Teevan, J. Searching to
eliminate
personal
information
management.
Communications of the ACM 49, ACM Press (2006),
58-64.
25. Karger, D.R. Haystack: Per-User Information
Environments Based on Semistructured Data. In Beyond
the desktop metaphor : designing integrated digital
work environments, MIT Press (2007), 49-99.
12. Dittrich, J., Salles, M.A.V., Kossmann, D. and Blunschi,
L. iMeMex: escapes from the personal information
jungle. In VLDB '05: Proceedings of the 31st
international conference on Very large data bases,
VLDB Endowment (2005), 1306-1309.
26. König, W.A. Referenzmodell und Machbarkeitsstudie
für ein neues Zoomable User Interface Paradigma.
Master Thesis, University of Konstanz, 2006.
27. Mandel, T. Windows vs. OS2, the GUI-OOUI war : the
designer's guide to human-computer interfaces. Van
Nostrand Reinhold, New York, USA, 1994.
13. Donelson, W.C. Spatial management of information. In
SIGGRAPH '78: Proceedings of the 5th annual
conference on Computer graphics and interactive
techniques, ACM Press (1978), 203-209.
28. Moran, T.P. and Zhai, S. Beyond the Desktop Metaphor
in Seven Dimensions. In Beyond the desktop metaphor :
designing integrated digital work environments, In
Beyond the desktop metaphor : designing integrated
digital work environments, MIT Press (2007), 335-354.
14. Fisher, D. and Nardi, B. Soylent and ContactMap: Tools
for Constructing the Social Workscape. In Beyond the
desktop metaphor : designing integrated digital work
environments, MIT Press (2007), 171-190.
15. Furnas, G.W. and Zhang, X. MuSE: a multiscale editor.
In UIST '98: Proceedings of the 11th annual ACM
symposium on User interface software and technology,
ACM Press (1998), 107-116.
29. Müller-Prove, M. and Ludolph, F. Dueling Interaction
Models of Personal-Computing and Web-Computing. In
Methodic and Didactic Challenges of the History of
Informatics (MEDICHI), Österreichische Computer
Gesellschaft (2007), 34-38.
16. Gerken, J. Orientierung und Navigation in zoombaren
Benutzungsschnittstellen unter besonderer Berücksichtigung kognitions-psychologischer Erkenntnisse. Master
Thesis, University of Konstanz, 2006.
30. Paal, S., Kammüller, R. and Freisleben, B. A CrossPlatform Application Environment for Nomadic
Desktop Computing. In Net.ObjectDays; Lecture Notes
in Computer Science, Springer (2004), 185-200.
17. Grün, C., Gerken, J., Jetter, H., König, W. and Reiterer,
H. MedioVis - A User-Centred Library Metadata
Browser. In Proceedings of the 9th European
Conference, ECDL, Research and Advanced Technology
for Digital Libraries, Springer (2005), 174-185.
31. Perlin, K. and Fox, D. Pad: an alternative approach to
the computer interface. In SIGGRAPH '93: Proceedings
of the 20th annual conference on Computer graphics
and interactive techniques, ACM Press (1993), 57-64.
32. Raskin, J. The humane interface: new directions for
designing interactive systems. Addison-Wesley,
Reading, Mass., USA, 2000.
18. Hutchins, E.L., Hollan, J.D. and Norman, D.A. Direct
Manipulation Interfaces. In User Centered System
Design: New Perspectives on Human-Computer
Interaction, Erlbaum (1986), 87-124.
33. Rexhausen, S., Demarmels M., Jetter, H., Heilig, M.,
Gerken, J., Reiterer, H. Blockbuster – A Visual Explorer
for Motion Picture Data. In INFOVIS '07: Proceedings
of the IEEE Symposium on Information Visualization
(INFOVIS'07), IEEE Computer Society (2007).
19. Jetter, H., Gerken, J., König, W., Grün, C. and Reiterer,
H. HyperGrid - Accessing Complex Information Spaces.
In People and Computers XIX - The Bigger Picture,
Proceedings of HCI 2005, Springer, 2005.
34. Seffah, A. and Javahery, H. Multiple user interfaces:
Cross-platform
applications
and
context-aware
interfaces. In Multiple User Interfaces: Cross-Platform
Applications and Context-Aware Interfaces, John Wiley
& Sons (2004), 11-26.
20. Jetter, H. Informationsarchitektur und Informationsvisualisierung für die Post-WIMP Ära. Master Thesis,
University of Konstanz, 2007.
21. Jetter, H. permaedia - Personal Nomadic Media for the
Coming Decade. http://hci.uni-konstanz.de/permaedia.
35. Shneiderman,
B.
Direct
manipulation
for
comprehensible, predictable and controllable user
interfaces. In IUI '97: Proceedings of the 2nd
international conference on Intelligent user interfaces,
ACM Press (1997), 33-39.
22. Jones, W. and Teevan J. Personal Information
Management. University of Washington Press, Seattle,
WA, USA, 2007.
23. Kaptelinin, V. and Czerwinski, M.P. Beyond the
desktop metaphor: designing integrated digital work
environments. MIT Press, Cambridge, Mass, USA,
2007.
36. The Deepamehta Community. DeepaMehta: Maintain
your head, and share it, http://www.deepamehta.de
9