|| Contents:
|| 1. Innovation: Patent Pending. [APPLICATION]
|| 2. Design: Window Management. [TECHNIQUE]
|| 3. Skills Transfer: Learning by Doing. [TRAINING]

= = = = =
* DEADLINE EXTENDED: You can still save $500 if you register online by 25 May for the June seminar in usage-centered design. Details at <http://foruse.com/seminars/>.
= = = = =
* Nearly 2000 leading professionals have subscribed to forUse.
Forward this issue to your colleagues. To join: <http://foruse.com/forms/subscribe.htm>.
Browse archives: <http://foruse.com/newsletter/>.
= = = = =

1. Patently Obvious.
Software and so-called business process patents are a subject that generates a great deal of debate. On the one hand, consider genuine innovations in user interface design that are original, unobvious, and broadly useful--precisely the sort of “inventions” for which patents were originally conceived, both to reward innovators and to spur further innovation. On the other hand, we find those increasingly common software and process patents that seem to fall in disputed territory where the ideas are obvious or the patent serves more to impede progress than to promote it. Many would argue that Amazon.com’s patent on “1-Click” purchasing represents proof of a patent process gone seriously astray.

The patent issue has been a recurrent one for us, because, aided by the abstract modeling techniques of usage-centered design, we have generated a steady stream of design innovations, a number of which have been considered for patent protection. Most of these “inventions” have been effectively placed into the public domain, either by virtue of our own public disclosure in courses and writing or because our clients chose to use them in products without initiating the patent process.

As a rule, we prefer to see good ideas put into practice and widely used. We gave away a number of our user interface innovations by including them in our book, Software for Use. These include the “ranked tabs animation” (p. 204-205), the “viewfinder graphical navigation scheme” (p. 205-206), “normalized context menus” (p. 179), “cascading tool tips” (p. 242), and “temporary tool trays” (p. 287). Whether or not any of these designs would have survived the rigors of the patenting process is moot; that they are novel and useful has been confirmed on a number of projects in which they have been successfully implemented. For instance, cascading tool tips, first introduced in this newsletter in issue #2, have been used effectively in several projects, most recently to great effect in the Step7lite programming system from Siemens AG.

We never set out to be inventors, but we were long ago convinced that to achieve real breakthroughs in usability, designers have to go beyond run-of-the-mill solutions based on standard widgets and conventional interaction idioms. Nevertheless, creativity for its own sake, especially of the artistic variety, all too often interferes with ease of learning and use. In our own work, we don’t strive to be different or original; we just try to produce the visual and interaction design that most closely fits the structure of the tasks, one that is easy for the targeted users to understand and efficient for them to use. We begin with the assumption that available components and traditional techniques of interaction will suffice, but when the limited repertoire of standard visual components gets in the way, we are not afraid to devise new schemes to solve particular problems.

The recent demise of a dot-com client left us with a fresh collection of orphaned ideas that had passed into the public domain when the client, ignoring the advice of its own patent lawyers, released the software without having filed patent applications. Not wanting to see all that work go to waste, we decided to turn some of the ideas into teaching material, hence the birth of our new series of Design Studies. The first study, reported in issue #13, described a combined content navigation and customization control <http://foruse.com/articles/designstudy1.htm>. The second one, on a specialized component that facilitates rapid selection among many alternatives through user-controlled highlighting, is now posted on our Web site <http://foruse.com/articles/designstudy2.htm> (The .PDF files are each about 220K).

We remain somewhat ambivalent on the matter of software patents. We believe inventors deserve to be compensated for their contributions to technological progress, but we are keenly aware that patents often merely feed corporate greed. Small businesses may be the most handicapped by the system as it works today, since the big guys all have their thick patent portfolios to cross-license with each other or to use to squish annoying upstarts. Is the answer for everybody to patent? Or should we all be giving it away and hope things just work out for the best? What do you think? <mailto:larry@forUse.com>

2. Washed-up Windows.
In the previous issue, we opened up the critical problem of designing the navigation scheme--the means by which users are able to move from one place or part to another within a Web site or application [“Navigating Uncharted Waters,” forUse #13]. Navigation can have a dramatic or even disastrous impact on ease of use.

Designers do not always think of window management or view management as an issue in designing application navigation, but switching among windows and managing them to keep visible what is of interest to the user is as much about getting around and keeping track of place as any other aspect of navigation. Window management can be a substantial part of the everyday overhead of using complex applications. It adds up when you consider the time a user spends resizing and moving windows around, using alt-tab or the Task Bar to try to get to the right window, plus moving modal dialog or message boxes or even floating tool palettes that always seem to block the needed view, and then include cascading, tiling, splitting, un-cascading, un-tiling, un-splitting, minimizing, maximizing, and normalizing of windows. Some unpublished studies suggest that window management can account for 8-10% of a user’s time. It is not only a significant source of inefficiency but a substantial contributor to user errors.

In three recent design projects, minimizing window management overhead has emerged as a major technical objective and, as it turns out, a powerful means for improving user productivity. The systems being designed were in wildly divergent areas of application with vastly different user communities: a browser-resident classroom information management system for K-12 teachers, a software development environment for automation programmers, and a medical records system for primary-care medical staff. These systems had in common that they placed a high premium on simple and efficient operation combined with strict limitations on available screen real estate.

The solutions devised by the design teams were unique to each application and its attendant details, but they also had some things in common. All three designs settled on single-window designs that would typically be run maximized. Full-screen operation is not a solution in itself, however, and simple screen-to-screen context switching can be both jarring and confusing. To achieve simple, rapid context switching among a large number of different views or application subsystems while maintaining the user’s sense of place and continuity required carefully worked-out combinations of techniques tailored to each application.

First, views and facilities were organized into collections or sets that were closely related and would be likely to be needed together. A tabbed notebook structure was used to organize each set, although the exact form of the tabs and their organization into a notebook or dialog differed with each design. In some cases, it turned out to be expedient to have a single view or facility actually “appear” in or be accessible within more than one set. In other words, the same “page” might be included in more than one “notebook.”

In two of the three projects, a scheme of multi-level or nested tabs was needed. For example, the classroom information management system as designed presented teachers with “folders” within “sections,” the former visually resembling conventional manila folders and the latter suggesting hanging file folders. The visual and conceptual metaphor of a set of labeled manila folders inside a tabbed hanging folder was completely familiar to teachers; the more conventional Explorer-style “tree-view” would have been easier to devise and much easier to program, but it also would have been less intuitable and far less effective with these users.

In all cases, familiar navigation elements were reworked to better fit specific purposes. For example, all three projects ended up with one of the main navigation elements being a vertical panel on the left of the screen. However, in two of these, this left navigation “channel” served as the primary or top-level navigation, allowing for the selection among any of a number of the available collections of tabbed views. For example, the Step7lite programming environment, has a columnar panel at the left as a way to select any several “view sets” comprising the parts of a project. Each view set consists of a several tabbed views that can be switched among. In contrast, the scheme that best fit the problem in another project used the left channel for navigation at the lowest level, taking the user directly to any of the many sections within a long record, while primary top-level navigation was provided by a set of tabs across the top.

Part of what made these somewhat sophisticated multi-level navigation schemes work was careful attention to behavior. For example, in these applications, the system is easier and more efficient for the user if the last page or tab chosen by the user remains “in-front” of that set when switching back and forth among sets. On the other hand, when starting a new project (or class or patient), the workflow is better if the each set reverts to its default order.

3. Do-it-yourself.
Want to invent better user interfaces? Come learn through hands-on training with the inventors of usage-centered design, Larry Constantine and Lucy Lockwood. Details: <http://foruse.com/seminars/>.

= = = =
forUse is published 9 times a year by Constantine & Lockwood, Ltd., trainers, consultants, and innovators in usage-centered design. On the Web at <http://foruse.com/>. © Copyright 2001, Constantine & Lockwood, Ltd.

To unsubscribe, send email with the word "unsubscribe" (no quotes) in the subject and message to: <mailto:unsubscribe@foruse.com>