Forum

The answer to the question "Is the Crouching Tiger a Threat?" in Robert L. Glass’s "Practical Programmer" column (Mar. 2006) is both yes and no. Yes in the sense that the U.S. position as computing/IT superpower is indeed threatened and no in the sense that more opportunity for rewarding employment in the world is good for everyone.

Any kind of work is a candidate for being performed somewhere else in the world. Increased use of robotics and telepresence means that even work requiring physical access to fixed objects could be affected.

The U.S. should view the emergence of Asia as a computing/IT power not as a threat but as an opportunity. Those of us in computing—where change is the one constant—should understand.

Bob Ellis
Fountain Hills, AZ

Robert L. Glass (Mar. 2006) wrote: "U.S. computer science academic programs are dominated by students from developing countries, especially Asian ones." Many of them stay in the U.S., some going on to found new U.S.-based enterprises that hire local technical talent. Who is threatened? Certainly not the students reflecting the "60% decline in the number of U.S. incoming college freshmen considering CS as a major." Many of them end up in business administration just so they can be hired by the foreigners creating the new enterprises.

The population of South and East Asia, stretching from India to Japan, is about 10 times that of North America. All things being equal, shouldn’t that population be expected to also have 10 times the number of computing professionals?

After taking an advanced degree in mathematics 35 years ago, I stumbled into programming, found it enjoyable, and have been with it ever since. During that time, the profession and its relationship with society has evolved. Just as we must constantly work on our technical education, sometimes to the point of reinventing our professional lives, we must also reevaluate our relationship with the wider world. We must realize the U.S. can no longer tell the rest of the world how it’s done. Saying so is not a threat. It just is.

Claud D. Price
Larkspur, CA

Robert L. Glass (Mar. 2006) asked whether the U.S. faces a global shift in IT competitiveness. To answer, consider that foreign nationals have represented the majority of CS students in the larger U.S. state university programs, and the terrorist attacks of 9/11 and resulting U.S. xenophobia and tightened immigration laws have contributed to the decline in the related number of CS graduates. Meanwhile, the verbal and written communication skills of many of these people limit their employability in major U.S. integrators—the main employers of CS and engineering graduates. This isn’t only a U.S. issue. Surveys show that many graduates of Indian and Chinese universities are not employable by multinationals in their native countries.

Demand for U.S. IT positions exceeds the levels of 2000 and is growing. India forecasts a shortfall of 235,000 IT professionals by 2008, and China forecasts a shortfall of 690,000 graduates (all degrees). Indian IT has boomed as offshore outsourcing has boomed. Competition is intense for skilled developers as salaries escalate. Publicly traded IT company valuations have risen significantly. India looks much like the U.S. did in 2000 when CS degrees and IT careers were generally considered attractive. Hot IT areas today focus on business process outsourcing. But markets don’t necessarily maintain their momentum, and this one is as likely to slump as the earlier ones.

Governments contribute best to their IT industries through education. In this sense, the U.S. leads India and China; 60% of U.S. students have attended some college compared to 13% in China and 7% in India.

Along with a crouching tiger, the hidden dragon in the U.S. IT industry trumps all these trends. Software innovation is driven by meeting the unmet, often unknown, needs of businesses and consumers. U.S. IT businesses also innovate software and services to serve emerging business models, like those used by Google, eBay, and Amazon.com. MySpace and the Face Book are U.S. phenomena. U.S. innovation will continue unless Asian companies that use software become business model innovators as well.

Another source of U.S. innovation is within IT itself. Like the Japanese in manufacturing, India has adapted the best U.S. IT practices to build an outsourcing industry around Capability Maturity Model Integration. However, the latest wave of best practice around agile methods using small teams in close collaboration with clients is relatively difficult for the Indian IT industry to embrace. It will first have to abandon its successful business models or overcome time and distance issues through improved virtual reality and collaboration technologies.

Regardless of the outcome, the U.S. will be better off if other countries close the IT gap, ultimately contributing to a better and more peaceful world.

Bruce Ballengee
Dallas, TX

Responding to Robert L. Glass (Mar. 2006), I’d like to point out that many Asian students enrolled in U.S. universities will eventually become U.S. residents or citizens and that many top graduates from Asian universities, particularly those in China and India, want to come to the U.S. to work for U.S. companies. U.S. computing/IT dominance is not affected in this regard. Meanwhile, as the world becomes more diversified, other countries are likely to see the IT industry flourish at home. It’s not necessarily a good thing for one country or region to dominate.

Jack Yang
Fresno, CA

Robert L. Glass (Mar. 2006) wrote: "The U.S. has dominated the computing profession for more than five decades now. It’s hard for my common sense to accept the possibility that we’re about to reach the point where that U.S. dominance is going to be overthrown." Then again … imagine it’s 1970. Substitute "automobile industry" for "computing profession" and apply the common sense test. To imagine more, read Thomas L. Friedman’s books, especially The Lexus and the Olive Tree and The World Is Flat.

Heman Robinson
Cary, NC

Back to Top

Give Metadata Its Due

As an industry practitioner who has built several metadata repositories for Fortune 100 companies, I was surprised that "The Metadata Enigma" by Ganesan Shankaranarayanan and Adir Even (Feb. 2006) included three inaccuracies:

Metadata standards. The metadata standards war is long over. The Microsoft-led Meta Data Coalition Open Information Model effort was folded into the Object Management Group’s (OMG’s) Common Warehouse Metamodel (CWM) in September 2000 (see The Common Warehouse Metamodel, John Wiley & Sons, Inc., 2002). Shankaranarayanan’s and Even’s article is six years out of date and did a disservice to much hard work and compromise in the standards community;

OMG. Oracle does not lead the OMG (as the article might seem to imply), and the CWM effort included lead authors from IBM, Hyperion, and Unisys;

Metadata integration. It is untrue to say that "a comprehensive commercial solution for full-fledged metadata integration does not exist"; several enterprise-class solutions are available today. The primary issue is making a business case for comprehensive (and expensive) implementation.

The article also included three notable weaknesses:

Process. Process represents an unaddressed but critical issue; what processes are needed to define and maintain metadata?;

Convergence. Technical metadata dovetails with the concept of the Configuration Management Database, as advocated by the IT Infrastructure Library specification, and the practice of IT service management, a convergence that needs further research; and

Regulatory drivers. Regulatory drivers (such as the Sarbanes-Oxley Act, the Health Insurance Portability and Accountability Act, and customer privacy) are arguably the most important recent trends supporting a business case for metadata investment. The article ignored them.

Charles Betz
Minneapolis, MN

Authors Respond:

We thank Betz for his insightful comments. He would likely acknowledge the difficulty of covering metadata in its entirety in a magazine article. Ours reflected the metadata-research perspectives that intrigued us in 2003 and 2004. Important research areas (such as the implementation process, the Configuration Management Database, and regulatory drivers) deserve further discussion. Failing to mention the merger of the two competing standards was an oversight on our part. However, the software tools we researched (in 2003 and 2004) did address the standards, and our objective was to highlight the incompatibilities that existed then. We are aware that companies other than Oracle participated in development of CWM, though Oracle did play a key role in the effort. Regarding metadata integration, no commercial tool we reviewed supported the comprehensive integration of all metadata categories, particularly process and quality metadata. Moreover, we share Betz’s concern about the difficulty of justifying enterprise-class solutions and are eager to learn about any positive progress in that direction.

Ganesan Shankaranarayanan
Adir Even
Boston, MA

Back to Top

Who Do You Trust?

We can apply the criticisms concerning Wikipedia by Peter Denning et al. in their "Inside Risks" column ("Wikipedia Risks," Dec. 2005) to the column itself. "You cannot be sure which information is accurate and which is not," they wrote. This is epistemologically incontrovertible but applies to any information, no matter how august its source. And, given that they made unfavorable comparisons between Wikipedia and the incontrovertibly august Encyclopædia Britannica, it is a shame that the study reported by Jim Giles in Nature ("Internet Encyclopedias Go Head to Head," Dec. 15, 2005) was too late for inclusion. That study found "… numerous errors in both encyclopedias, but among 42 entries tested, the difference in accuracy was not particularly great: the average science entry in Wikipedia contained about four inaccuracies; Britannica, about three."

Denning et al. also wrote: "[Wikipedia] cannot attain the status of a true encyclopedia without more formal content-inclusion and expert-review procedures." Given the failure of editing and peer review in some recent cases, it is possible that a large number of self-appointed experts, amateurs, dilettantes, and informed lay people may well produce a more accurate encyclopedia than a small number of the officially approved. It is certainly more fun.

Adrian Bowyer
Bath, U.K.

Authors Respond:

Since we wrote the column, media accounts have featured people purposely falsifying Wikipedia articles and U.S. Congressional staffers editing articles about their bosses to remove unflattering facts. They illustrate our concerns.

The Wikipedia study reported in Nature did not allay them. It involved a tiny, statistically insignificant sample in a narrow range of topics at a single point in time. As we noted, Wikipedia’s quality varies with subject matter, and its anonymous mutability renders snapshots meaningless. It is not a formula for a reliable reference work.

This exchange with Bowyer illustrates one of our points: You know who stands behind our "Inside Risks" column and who stands behind Bowyer’s comment; you don’t know who stands behind most articles in Wikipedia.

Peter Denning
James Horning
David Parnas
Lauren Weinstein
The authors are members of the ACM Committee on Computers and Public Policy

Back to Top