GUEST EDITOR’S INTRODUCTION

User Interfaces: Opening a Y, Deborah Mix, Virginia Polytechnic hstitute and State University

I.‘ z:‘:” evolutions bring change - ” ii : often change for the better. & q~,G ” :‘ Computer science has wit-

nessed many revolutions, including the emergence of high-level progl-am- ming languages, database-manage- ment svstems. and artificial intel- ligcnce.‘Each in itsown way demanded changes in how software is developed. Now, the advent of interactive systems with human-computer interfaces is causing a revolution thatdemandsnew approaches to developing software of a different sort.

Building any software system is a complex tisk, even when the system is complltation-intensive and does not cmphasile the interface. Injecting the ever-changing, always-unsatisfied human into the process necessitates a quality human-computer interface, complicating an already complicated task. An interactive system -one with a human-computer interface - is not

judged solely on its ability to compute. It is alsojudged on its ability to commtl- nicate. In fact, if users cannot commu- nicateeffectivelywith an interactive sy.s- tern, itr computational ability may be inaccessible.

Developing an interface is very different from developing conven- tional software. The traditional meth- ods, techniques, and tools that work well for software development are proving not to work so well for inter- face software. How many softwarede- velopment methodologies specifically include procedures for producing the interface? New technology is con- stantly increasing the developers’ palette of interaction devices, styles, and techniques, yet how many conven- tional languages have constructs for in- corporating a window, an icon, or a mouse?

We must support the revolution in userinterfaceswithnewmethods, tech- niques, and toots. Our goat, of course, is better interfaces.

Articles. The articles in this issue ad- dress the revolution’s mo areas, from both a theoretical and

8

pragmatic p&spective. Virtually all the ar- ticlesaddress what is perhaps the most far-

of the revolution: the ad- #ve systems for developing

rface. Thus this issue gives a multi-

074@7459/89/4JcO1/0008~0l.~@ 19R9 IkxE

faceted look at approaches to developing interface SOfhVare of a different sort:

l The first article, by Brad Myers, is an overview of interface-development sys- tems, which he calls user-interface devel-

IEEE Software

macro language. Kasik and colleagues describe the lessons they have learned from both users and developers.

Another research area in interface design iscommunication issues, especially how application semantics fit into design of the interface. Developers advocate sep aration of the interface from the applica- tion, but where and how to achieve that separation is not easy to determine, espe- cially for direct-manipulation, highly in- teractive systems. Two articles address this problem:

l First, Rex Hartson addresses how to separate components in an interactive sys- tem, illustrating and comparing ap- proaches to separation. He then describes how to achieve control and communi- cation among thosecomponents. Hartson presents the concept of two communi- cation levels to support separation.

l Next, Dave Hurley and John Sibert de- scribe how to model the interface between the user interface and the application. Their conceptual modeling system con- tains constructs and abstraction mecha- nisms that structure knowledge in amodel and primitives that give behavior to knowl- edge in a model. They use the interaction semantics of a spreadsheet to illustrate the practical application of their system.

l Once we learn how to produce a user interface -especially by using interactive tools - what do we do with it? The final product, the interface itself, must then be

evaluated. John Thonras and Wendy Kel- logg explore an ecological gap between what is observed during interface evalua- tion in a laboratory and what happens when the interface is used in the real world. Thomas and Kellogg look at the ecological gap as it relates to users, tasks, systems, and the real-world context. The) propose specific techniques to be used at different phases of development for eval- uation and tuning the interfaces in a real- world context.

This article is a nice closing for this special issue because it points to the need to do more empirical work on the human- computer interface development process. It is particularly appropriate for this publi- cation; publishing such articles primaril) in human factors journals propagates the very gap Thomas and Kellogg address!

Acknowledgments I thank Ted Lewis for his enco~rage~nent

and patience in guiding me through the task of editing this special issue, and June Martin of his staff. Thanks also to those who pl-ovidrd timely and thoughtful reviews; your efforts vastly improved the quality of the final prod- uct.

This issue would not have happened without .Joanne I.ee Bogner, my secretal-?, who kept track of the motuntain of details, kept authors (and me) on schedule. handled tons of’ror~-e- spondence, fotcnd answer-s to unanswerable

questions, and on and on. Rex Hartson and Bob Schulman gave much moral support and encotlragernent. Here it is, guys!

Deborah Hi is a research computer scientist on the faculv at \‘irginia Polytechnic and State University in Blacksburg. She is a principal in- vestigator on the Dialogue Management Prqj- ect. This project is concerned with how to achieve quality human-computer interfxes through development of concepts for human- computerinterfacemanagementandthrough development of specialized methodologies, techniques, and tools for producing the human-computer intuf’ace.

Hix received a BS in mathematics from Emory University and an MS in information scienceandaPhDincomputerscienceandap plications from Virginia Polytechnic Institute and State University. She is a member of IEEE, ACM, and the Human Factors Society.

Address questions about this issue to Hix at Virginia Polytechnic Institute and State Uni- versity, Computer Science Dept., 562 McBryde Hall, Blacksburg, \‘A 24061.

he bc IW wt

k r e

Get a free video from the ” leader in CASE.

It’s called “The ExceleratoP Difference? And it’s a frank, management-

o-management discussion about why Excelerator from Index Technology is so

different from other CASE solutions. And why Excelerator/RTS’ superior capabilities-with

Index Technology’s implementation support-can make a difference in your organization.

In this short video, you’ll learn how Index Technology has helped the leaders in aerospace, defense and engineering. And how we can help you