 |
 |
forUse Newsletter Archives | |||||
 |
 |
||||||
 |
|||||||
|
|
|
forUse Newsletter Archives | |||||
|
|
|
||||||
|
|
|||||||
 |
forUse: The Electronic Newsletter of Usage-Centered Design #27 | December 2002 |
|
 |
||
| Subscribe! |
   |
|
|
Contents: || 1. Conference: Calling All Colleagues || 3. Resources: From RUP to XP and Constantine to Wirfs-Brock
= = = = = 1. Conference: Second Annual Gathering If you missed the exciting and informative gathering of the clan this past August, you get another chance. After the unqualified success of the first conference, we agreed to organize the Second International Conference on Usage-Centered Design to be held 19-22 October 2003 in Portsmouth, NH. The theme, “Performance by Design,” reflects the usage-centered design focus on enhancing user performance through better user interfaces tailored to the tasks. We are psyched about the program, which is already shaping up. Bill Buxton, Chief Scientist at Alias|Wavefront, will keynote the conference. Buxton is well known as an innovative designer who challenges assumptions about what works and why. His work reflects a particular interest in the use of technology to support creative activities. We are also putting together a really exciting event, a panel titled “Between Extreme and Unified,” that will feature two teams of experts addressing user and usability issues across the spectrum of modern development practices. CALL FOR PARTICIPATION. 19-22 October 2003. (And please note that forUSE 2003 is NOT the same week as OOPSLA 2003 in Anaheim, which was incorrectly posted on the OOPSLA Web site.) 2. Design: Naked Objects Amok on the Interface. If you do not already know about Naked Objects, it is
probably time to learn. It may sound like the jocular name of a raunchy
hard-rock band, but this is an approach to be taken seriously by all
usability and user interface design professionals. The roll-out of Naked
Objects (the book) and Naked Objects (the programming tool) took place at
the recent OOPSLA Conference in Seattle. In a curtained off area of the
exhibit hall, Richard Pawson and Robert Mathews of CSC’s Research Services
in the UK demoed their Naked Object framework and preached the Naked Object
approach to user interface design. Their solution for usability? Eliminate
user interface design altogether. In its bare essentials, the Naked Objects premise is
simplicity itself. Instead of designing, building, and testing a user
interface for the next software system, all developers need to do is create
software objects that correspond to and fully model the “business objects”
that make up the application domain in the real world. The Naked Object
“framework” then automatically generates an “object-oriented user interface”
that simply presents these objects and their related operations (“methods”
to the object-oriented cognoscenti) directly to the user without adornment
or embellishment. The software objects are, thus, stark naked to the user. The user of an application--any Naked Objects
application--sees a series of windows containing icons representing classes
of objects or individual object instances. Objects can be “opened” by
double-clicking to reveal their contents in a standard format that lists
attributes of the object and their current values. To do something with the
system, users either drag-and-drop one object onto another or right-click on
an object to pop up a context menu from which operations supported by the
object can be selected. That’s it. All interfaces look essentially the same
and work the same way. What’s wrong with this approach? What’s right with it?
And why should you care? The usability problems with Naked Object interfaces are
fairly obvious. Context menus are a form of hidden behavior. Their contents
change and must be perused on every use in order to pick an action. Both
drag-and-drop and right-clicking are interaction idioms that a significant
fraction of users find difficult to master. With everything in separate
windows, window management can become a burdensome overhead. Naked Objects
also force users always to pick an object, then an action, even when it
might be easier or more natural to pick an action and then the object on
which to perform it. Perhaps the biggest problem is that the presentation of
information and the operation of the user interface are not tailored to fit
the unique aspects of the application and user needs. One solution is
presumed to fit all problems, provided all the relevant domain objects are
properly identified with all their important behavior fully modeled. What is right about the approach is the accompanying
rhetoric. Pawson and Mathews claim to be on the side of users and to empower
them by delivering “expressive systems” (the original and less sexy name for
Naked Objects). Expressive systems support creative problem solving rather
than imposing a strict order and logic on users as do so-called “optimized
interfaces” (the fantasy bogeyman of user interface design that Pawson and
Mathews erect and then attack with vigor). Pawson and Mathews are on firmest ground when they argue
for collaborative modeling with users to develop an accurate and complete
domain model. Moreover, with Naked Objects, they can promise to deliver a
system and a user interface that precisely reflect and fully support this
domain model, since the software objects and the interface objects are just
the domain objects in the model. You should be interested in Naked Objects not so much for
any innate viability as a visual and interaction design approach or as a
means to simplify or eliminate user interface design. You should be
interested because Naked Objects are likely to catch on big-time with two
groups: managers and developers. Managers and business decision makers will
embrace Naked Objects because they promise to cut the cost of usability and
user interface design to zero. The user interface--the “correct” user
interface--comes free, generated automatically by a free, open-source
software framework. No need to waste time on usability inspections or
testing, no need to pay interaction designers; just program the software
objects, crank the code through the open-source software, and your user
interface is finished. Programmers and software engineers, who have never
been completely comfortable with usability people in the first place, will
also love Naked Objects because the approach proves you don’t need user
interface designers or usability testers, who just get in the way of
delivering on time anyway. The straw man Pawson and Mathews construct in the form of
so-called “optimized interfaces” may be a stereotype, but it is an
uncomfortably familiar one. All too often, naďve designers with simplistic
notions of task-centered design violate the fundamental rule of user control
and produce systems that enforce an artificially rigid workflow. Many users
today feel overpowered by Microsoft-style adaptive interfaces that wrest
control from them by changing menus at will or interrupting willy-nilly to
offer suggestions or help that is never helpful. Many have grown tired of
so-called wizards that march them blindly down unmarked paths with an
implicit I-know-how-and-you-don’t message. Readers of this newsletter know that the stereotype of
such authoritarian interfaces is out of step with usage-centered design as
conceived and as practiced. Pawson and Mathews seem unaware that
usage-centered design can deliver user interfaces that are custom-tailored
to the task structure and the mental maps of users while also supporting
flexible problem solving. (I drew their attention to the Siemens STEP 7 Lite
system (www.foruse.com/pcd/)
as an example of “optimized” design that left users firmly and flexibly in
control.) Pawson and Mathews claim that we user interface designers
are threatened by their approach because we have a vested interest in
maintaining design as a costly black art requiring long apprenticeships. Of
course, they and their fellow object modelers also have a vested interest in
trivializing user interface design. Nevertheless, we should take the threat
of Naked Objects seriously and take the message of Pawson and Mathews to
heart. The lessons to be learned--or relearned--are that we need
to work closely with and listen to users to understand the true nature of
their work and the problems they face. Then we need to reflect that
understanding in the systems we design and build. We need to treat users as
intelligent problem-solvers by producing innately flexible designs that
accommodate to varied styles of work and problem solving, by creating
interfaces that do not enforce a rigid logic or artificial order and that do
not deprive users of the opportunity to devise better ways to achieve their
own ends. This is, of course, precisely how usage-centered design
works when practiced as intended. By thoroughly modeling the relationships
between users and the system, the work to be accomplished, and the content
to be supplied, we can give users better tools to accomplish diverse ends by
diverse means. Naked Objects may not be capable of delivering such tools,
but you can bet that their seductive message will have broad appeal. (For a more detailed analysis of Naked Objects, see the full article at www.foruse.com/publications/. For more on usage-centered design in general, see www.foruse.com. For more about Naked Objects, see www.nakedobjects.org.) 3. Resources: From RUP to XP and Constantine to Wirfs-Brock Learn about the leading edge in usage-centered design,
performance-centered design, scenarios and customer stories, agile design
and more. Get your copy of the forUSE 2002 Conference Package available now
at a special discount price. Gloria Gery, Karen Holtzblatt, Jim Hobart, Rebecca Wirfs-Brock,
and Jeff Patton are among the recognized experts whose important papers and
presentations are included in the Conference Proceedings and CD, which cover
a wide range of vital current topics, including: Usage-Centered Design in XP and Agile Development Usage-Centered Design and the Rational Unified Process Designing Information Portals Best Practices in Design for the Web Performance Support for Complex Problem-Solving Tasks Usage-Centered Exploration to Speed the Initial Design Process The Role of Scenarios in Usage-Centered Design Web Design Patterns for Transactional Systems What It Really Takes to Handle Exceptions in Use Cases Designers as Change Agents The complete package (www.foruse.com/2002/package.jpg)
includes: * Conference Proceedings – a handsome, professionally edited, 484-page perfect-bound book with 36 chapters. * Conference CD-ROM – Presentations in PowerPoint and Acrobat, plus a searchable electronic copy of the Proceedings; over 239 megabytes in all. * Session Handouts – Bound copy of all hardcopy session handouts. * Bonus (while supplies last) – large, conference
carry-bag to keep it all in. To get details and order, click here. =
= = = To unsubscribe, just click here to send email. |
||
| Subscribe! |
|
|
|
|
||