This year’s LinuxWorld West was was smaller than last year’s. However, according to the folks at IDG, the organizers, there was a 10% increase in overall attendance. This is in spit of the fact, as shown in Figure #1, that by most measures the advertisements, sponsors, tutorials and technical sessions were fewer this year. In discussions with the exhibitors on the show floor, I was told that the business quality of the attendees was better than in previous years. This lends credence to the idea that Linux is gaining commercial acceptance.

I also believe that IBM is commited to Linux. From their viewpoint, it is the unifying O/S of the future capable of running on all their systems in a uniform fashion. Can you image the cost savings to IBM? They never made a profit on all their operating systems.

There were two interesting new hardware products that I saw at the show.

The ED.1 from Vercel (www.vercel.com) was a triple combination of MP3 player, when folded up. A gameboy console and player when opened up, and a full-fledged computer and keyboard running Linux when the integrated keyboard was unfolded. It all fitted into a form factor that was about 5” x 4” x 2” and weighed less than 12 ounces. The unit was able to send and receive email and browse the Internet. Using software from Trolltech and others, it was also a full-featured PDA. Picture #1 does a poor job of showing its feature set and ergonomic design.

The second piece of hardware that caught my eye was the Panoram display (www.panoramtech.com). The unit consisted of three flat segmented TFT/LCD displays that were integrated into one unit. Its megapixel resolutions allowed for an immersion experience that I have not felt before. The effective screen size was 36”

Playing games on this unit would be a real trip.

There were few software surprises. The show emphasized enterprise management and security.

The major players were there. IBM promoted its developerWorks (ibm.com/developerWorks/linux). Sun touted their Sun Open Net Environment (Sun ONE), and Microsoft was talking about their idea of interoperability. Also, according to the signs at the Sun pavilion, each attendee at the Sun’ Network 2002 Conference (www.sun.com/sunnetwork) is eligible for one of Sun’s new desktop computers for the price of admission to the conference.

The lice debugger from LTrix Engineering (www.ltrix.com) impressed me. It is a GUI based low level debugger that can help resolve problems in kernel and driver code. They say it works without source code changes.

At the IBM Presentation, Jiang Guang Zhi, from Beijing, discussed “Why Linux?” Among his more telling points was that Linux was the proof of a new concept in software development, and that there are enough skilled Chinese to have a major impact on software development. I also think that they expect to expand rapidly and they did not want to be hampered by proprietary operating systems.

As discussed in other parts of this itssue, I attended Linuxworld West. It was a “ball”. I was there all three days and still did not get to see and here everything I wanted. Even though the show was smaller than last year, the level of enthusiasm was much higher than last year. The determination to see Linux succeed was great and with the large players really making commitments to Linux, the combined effort was starting to pay off. Even the “suits” were extolling the benefits of Linux and its credence in the corporate environment is established.

There was also a shake out—most of the smaller Linux distributors were gone with some of them combining into United Linux. For example, I was told that the development staff from Caldera are now working on United Linux, but are being paid by SuSE.

Based on my experience this year, I am really excited by what is coming next year.

Book Reviews

Software for Use: A Practical Guide to the Models and Methods of Usage-Centered Design, by Larry L. Constantine and Lucy A.D. Lockwood, Addison-Wesley, ISBN 0-201-92478-1

Software for Use presents a process for creating better, more usable software. The book is aimed at developers, but is useful for anyone involved in development. The focus is mostly on desktop applications, but one chapter is devoted to issues unique to the web and embedded applications. Most of the book describes creating a new application, but the last few chapters discuss methods for conducting inspections and identifying user interface issues in completed applications.

Usage-centered design is an iterative, interactive process. The authors emphasize good design is not an accident – it is an important, visible part of the project from the start. After gathering user requirements, you should create a list of user roles: a name and description of who will use the software. The heart of the process is determining what the users actually want to achieve. From there, you generate essential narratives describing user intentions and system responsibility for each step of each use case. Using this data, you create an abstract prototype, a navigation map, and finally, a visual design.

This book gives a structured process for what you need to consider and how to generate and test hypotheses. Several means for generating essential use cases should be familiar from OOD: inheritance, extension, and composition. Describing the process in these terms makes it accessible to developers. Also. with a single example throughout the process description, it is easy to see process components building on each other

As you may have gathered, there is a lot of jargon in usage-centered design. The authors define their own terminology and differentiate it from other, previous work. This slows down the early chapters. A developer may have difficulty convincing a project manager that spending time writing essential narratives will be worthwhile, so it would help to include more information about why a good user interface is valuable.

Whatever components of usage-centered design you take from this book, the message is to think carefully about how people will actually use the software you create.

— Lutz Birkhahn

The overall job market continues to stagnate here in the Valley. Software Engineering job openings shrank by 4%. Looking at the major platforms that the market currently supports, only Windows ME and Windows XP showed some growth since October of last year. Out of about 1000 job openings currently listed, 260 involve Windows ME and 30 involve Windows XP. This is shown in Figure #1.

In Figure #2, we have rationalized the job data to show relative strength of each platform. That is, the number of jobs for a given platform relative to the total number of jobs available. Thus, Windows XP only represents 3% of the jobs, but it, Windows ME and Windows 2000 are the only ones showing an increased demand for software engineers with those skills.