Workshop reviews

Human Centric Computing Languages and Environments (HCC) and Empirical Studies of Programmers (ESP)

2-6 September 2002, Arlington, VA

By Marian Petre

The good news is that 'ESP is back' - this time as one of three symposia at Human Centric Computing Languages and Environments (formerly Visual Languages), an IEEE-sponsored conference.

HCC is organised into distinct symposia, each with its own, largely autonomous committee. (Susan Weidenbeck, of Drexel University, and I chaired this year's ESP.) Submissions to ESP were strong, and there were good interactions with the other symposia: Visual/Multimedia Programming and Software Engineering, and End-User Programming.

Although HCC is suffering - like most US conferences - from lowered attendance, the conference was vigorous and interesting, with international attendance of about 60, including familiar faces from both the ESP and PPIG communities.

HCC featured three provocative keynote talks, all of potential interest to PPIG members:

  • Ben Bederson, Human-Computer Interaction Lab, University of Maryland:
    Interfaces for staying in the flow

Ben introduced interface design guidelines relating to the principle of encouraging 'flow', or 'optimal experiences' in interactions. The gist was to minimise cognitive load while maximising subjective satisfaction.

He of course showed some nifty examples, such as 'flow menus' (Guimbretiere et al.) and 'photo menus' (on his web site). And there was some interesting discussion about how 'flow' relates to different sorts of experiences: team working, generative tasks, and responsive/reactive tasks.

  • Randy Pausch, Carnegie Mellon University:
    Alice - a system using 3D graphics to teach computer programming

Alice is a largely drag-and-drop programming environment for animating 3D models, aimed at 'middle school' students. It has undergone years of development and a variety of evaluations, and it has sex appeal.

It's hard to beat Randy Pausch as a quotation generator (and quoter), e.g.: 'Code continues to be the best way to express program logic, even for novices'., 'If' statements aren't that hard; the syntax for anything is hard...the intellectual effort in a typical intro programming course is more than half teaching syntax to a particular language.', 'Parallel programming is child's play; synchronisation is hard.'

  • Clayton Lewis, University of Colorado at Boulder:
    HCI - where are we now, where might we go from here?

Clayton is, deservedly, a PPIG hero. He presented a survey of 'influential talks' and key developments in HCI, and he considered 'What now?' Recurrent themes included that 'We've learned that empirical, feedback-driven methods will produce useful and usable systems' and 'The critical issue isn't technology, it's engagement with the details of the problems' and 'Engaging the specifics requires crossing discipline lines.'

He made an appeal that researchers engage the underlying, human problems - the problems behind many of the problems we study. He predicted that within 10 years 'advances in the science of human nature will contribute in fundamental ways to the work we do.'

Susan Wiedenbeck and I led a panel/discussion session on Empirical studies of programming-in-the-large which generated considerable energy and interest. This was a reprise of a similar session at IWPC a couple of years ago and drew in results from a highly-informal PPIG survey at that time.

Its focal questions were: What are the questions that can only be addressed by studying programming-in-the-large? and What are the big questions in software engineering / programming that should be addressed by empirical research? We expect to collaborate on a paper based on those discussions.

My most memorable image from the conference is from Randy Pausch: a photo of one of his programmers doing user testing sitting on his hands. Their rule is that the programmer can't help the users without admitting: I'm sorry I wrote such a crappy system; let me help you overcome it.

The Proceedings are published by IEEE, ISBN 0-7695-1644-0.