Forum

I agree with David A. Patterson’s "President’s Letter" ("Computer Science Education in the 21st Century," Mar. 2006) that for CS to be interesting to students, it must also be inherently challenging and relevant. I have too often had to turn away job applicants because they wanted to sit undisturbed writing a masterpiece from scratch. Industry has no time for that. It needs requirements collected and analyzed, user documentation written, code tested, existing code adapted, and occasionally new code written but only as glue for preexisting blocks.

In the lithography community where I come from, we’d all be fired if we delayed a billion-dollar fab because somebody decided to field some "bright idea" in software. It may be somewhat boring, but good software engineering is no different from any other engineering specialty in this regard.

Patterson defined the root cause of the poor fit between today’s CS graduates and the needs of industry, writing "There is a huge disconnect between the experience of most professors, who have never worked as professional programmers … and the way in which cutting-edge software is written today." Perhaps the profession would help itself eliminate the inbreeding sometimes seen in universities, insisting instead that new faculty be drawn exclusively from industry.

Richard L. Lozes
Dublin, CA

In a recent effort to recruit interns and new grads at Apple Computer, I found the number of them familiar with parallel programming issues (such as threading and locking) was rather low compared to those familiar with, say, popular CS algorithms. While recognizing that algorithms and data structures are important, the 21st century, as David A. Patterson pointed out (Mar. 2006), is going to be a tale of reworking virtually all software to perform more efficiently in parallel. We need CS graduates who understand these techniques.

Andy Belk
Menlo Park, CA

Back to Top

The Tiger Means Healthy Competition

In his "Practical Programmer" column ("Is the Crouching Tiger a Threat?," Mar. 2006), Robert L. Glass raised five specific concerns. None of them, however, will turn out to be a serious threat to the U.S. position in computer technology. First, quantity does not equal quality. The combined population of China and India is nearly 20 times the population of the U.S.; meanwhile Japan’s population is almost 128 million and Indonesia’s almost 246 million. When more families in these countries have enough money to send their children to college, whether at home or abroad, the number of Asian students enrolled in CS (and everything else) will be greater than the number of native U.S. enrollment. Similarly, there are ultimately likely to be more IT practitioners in Asia than in the U.S., along with more IT-related research papers written and published by Asian institutions.

In terms of quality, Asian computing still has a long way to go. How many top CS scholars and researchers today are from the Asian computing industry? And how many top Asian software companies have their own branded products—the equivalent of Windows, Oracle, and SAP—that dominate a particular software market?

Glass was concerned that "U.S. dominance is going to be overthrown." Even if Asian computing would soon challenge the U.S.’s dominant position, it represents healthy competition and the promise of a better computing community for everyone worldwide, not just in the U.S.

U.S. success is largely the result of a combination of the country’s scholarship and practice. Both are motivated by the challenge posed by other countries, including those in Asia, trying to improve their own scholarship and practice, not because the U.S. has figured out a way to "get rid of threats."

Finally, why didn’t Glass express similar concern about the threat posed by the European computing community against U.S. computing dominance?

Wayne Huang
Athens, OH

Robert L. Glass (Mar. 2006) wrote that the U.S. has dominated the IT industry thanks to an ecosystem involving industry, universities, and government. But today U.S. government investment in IT research may be falling off, leading to reduced interest by students in university CS programs. The ecosystem may finally be losing energy on its own.

U.S. IT dominance is characterized by two main factors: The first is U.S. companies, which lead because they are able to tap talent regardless of geographical location; relatively few non-U.S. companies produce equivalent world-class software products. The second is programmers, but offshoring is likely to erode U.S. dominance.

The real danger to the U.S. is not the loss of IT jobs but deflation of the ecosystem. Software leaders in U.S. universities and the U.S. government must therefore keep funding the research. When funding is available, students follow. Meanwhile, business leaders are less likely to focus on the U.S. software industry, as their cost-benefit analyses no longer reflect U.S. programmer dominance.

Umesh Panchaksharaiah
Richmond, CA

As you may guess from my name, I’m originally from one of the countries (Korea) cited by Robert L. Glass (Mar. 2006). I’d like to share what I think are the implications of the facts Glass explored. For example, he wrote about the increasing number of international students and about Asian research publications. But in fact the best international students stay in the U.S. after school and represent a serious "brain drain" for their home Asian countries. Even though many other students return home to teach and try to make their home universities more competitive, U.S. universities have maintained their dominance.

Some Asian universities evaluate academic performance based on number (not quality) of journal publications, but their researchers tend to publish more marginal-quality articles than what could be called high quality. I am an alumnus of the Korea Advanced Institute of Science and Technology but admit such universities are nowhere near the top 10 U.S. universities (at least for now).

The number of students majoring in CS is declining, even as the number of programmers is increasing in Asia. The reason in the U.S., as well as in some Asian countries (such as Korea), is that top students can earn much more by becoming lawyers or physicians than by becoming computer scientists. Bill Gates and the dot-com bubble made some people re-think this proposition, but their reflection didn’t last long.

In Asian countries, programming jobs are considered (relatively) desirable and stable. They are seen as much better than, say, sales/marketing jobs and may be a reason for the increasing numbers of professional software developers in Asian countries.

The real problem is globalization. If you (perhaps the majority in the U.S.) value globalization you would be unlikely to view globalization as a threat. Outsourcing programmer jobs is like buying made-in-China goods at Wal-Mart. Globalization delivers cheaper goods and cheaper software. Anyone unhappy about that should join me (and many others) in opposing globalization.

Zu Kim
Berkeley, CA

The Asian tiger is indeed trying to take over the IT world. Apply the common-sense test to radios, TVs, cameras, and cars. How long did Europe and the U.S. dominate the world of RF communications, CRTs, photography, and automotive engineering, not to mention DRAMs and LCDs? And how quickly did they lose their hegemony?

Wait, you might say, these are manufacturing issues. How exactly will software spare us from the same forces that moved dominant automotive engineering from General Motors to Toyota? Cost of manufacturing alone does not explain why the market for video game consoles isn’t dominated by U.S. vendors.

Meanwhile, leading scientists worldwide come to the U.S. after being forced to retire back home, then produce the most advanced research of their lives. This is an interesting point and would hold sway if we were talking about scientific discovery. This counterargument didn’t do much for American Motors.

In The Rise and Fall of the Great Powers 1987], Paul Kennedy, a history professor at Yale University, wrote that a society’s commitment to education is a leading indicator of its economic and political power. Without a commitment to education, a dominant society (think England in the first half of the 20th century) is likely to see its influence in the world reduced to a level commensurate with its share of the world’s population and natural resources. The decline of U.S. hegemony in the IT arena appears inevitable.

Collin Park
Redwood City, CA

Back to Top

Shocked By Email Color

I was shocked by how the article "Does Color in Email Make a Difference?" by Moshe Zviran et al. (Apr. 2006) encouraged spam. It described an experiment in which the authors sent junk email to a mailing list of 1.4 million addresses and studied how the color of the email affected the number of responses. I’m not enough of a lawyer to know whether or not this experiment violates any laws or conventions about human-subject experiments, but it certainly does not seem like the type of work Communications should seem to be encouraging. It has value only to spammers and was done by enlisting the involuntary cooperation of 1.4 million people.

David Evans
Charlottesville, VA

Authors Respond:

While the Internet is often misused and many commercial email messages are spam in the sense they are unsolicited, inappropriate, and of extremely low value, the email in our study was certainly appropriate. It was sent to active users of IncrediMail who had voluntarily agreed to receive periodic email as part of IncrediMail’s particular business model of e-commerce. IncrediMail’s messages are functional messages that inform its users of new technological developments and system enhancements. Any medium can be misused, but the more we learn about media use, the better we are able to recommend how to do so in effective and just ways.

Moshe Zviran
Dov Te’eni
Yuval Gross
Tel Aviv, Israel

Back to Top

Pursue Reality As a Competitive Advantage

Amy S. Bruckman’s response to Curtis Rhodes’s "Forum" comment ("Be Skeptical of Rhetorical Slight of Hand," Mar. 2006) is even more objectionable than her "Viewpoint" "Student Research and the Internet" (Dec. 2005), which prompted Rhodes’s comment in the first place. She claimed "[An objective view of reality] is but one of many competing views about the nature of truth in a spectrum from objective to subjective."

Subjective views of reality are a luxury only stable or slowly changing societies can afford. That is because anything other than an objective view of reality always involves some amount of denial. In a quickly changing environment, denial is destructive to the society. People (and societies) who see past it to objective reality have a competitive advantage over those who hold to other ideas.

Bruckman does her students a grave disservice by suggesting that the many different views of reality are equivalent and can be selected like one selects articles of clothing to wear for the day.

David Randolph
Plano, TX

Author Responds:

Recognizing that multiple views of reality exist and other people’s epistemologies may differ from one’s own is essential for understanding others, whether you agree with them or not. However, the fact that the views exist does not imply that all views are equivalent. In our class at Georgia Tech, we teach students basic ethical theories (such as Kantianism, utilitarianism, and social contract), trying to prepare them to make ethical decisions as computing professionals.

Amy S. Bruckman
Atlanta

Back to Top