LETTERS: The Readers' Forum

Not all IT’s fault
I found Uche Ogbuji’s column (“Be humble, not imperial ”) in the December issue thought-provoking. It’s hard to argue with the characterization of the reputation of IT at most firms in the first half of the column. I do think, however, that the business side has contributed to the situation by being mostly as guilty of not learning about the capabilities and limitations of IT as IT is always accused of being regarding business knowledge.

As an IT professional, I do find it offensive whenever I read a business person stating his or her belief that it is easier to train a business person in technology than a technology person in business.

What most interested me, however, is that, without saying so, Mr. Ogbuji seemed to mean by “imperialistic data design” the fundamental principles that stem from the mathematical roots of the relational model. While I can’t believe he meant to suggest that heretofore accepted best practices in data modeling, such as normalization and logical/physical separation, are wrong-headed or obsolete, that seems to me the only logical conclusion to be derived from his statements.

I don’t know what the answer is. It is true that “imperialistic data design,” while certainly responsible for enabling the widespread, successful use of relational database technology we take for granted today, does often seem to lead to non-agile states of affairs. I think we are still a long way from being able to apply IT to support business agility with the reliability and routine that are taken for granted in other engineering fields. IT inherently deals in abstractions and has correspondingly much higher complexity and chaos. As long as the pace of business change remains as fast as it is, I think the application of IT to business support will always be sub-optimal by definition. —Mark Frawley [email protected]


Value-added analysts
I was glad to see an article on the relationship between business and IT in the December issue (“The IT-business relationship moves into a new phase ,” by Wayne Eckerson). I agree that there is a problem in many organizations because the two groups do not understand each other’s work. I read eagerly toward the solution that I was expecting, but was disappointed when Mr. Eckerson discounted the solution of the business analyst. Although the role does require a unique, knowledgeable individual, as Mr. Eckerson points out, these individuals are not rare. There are hundreds, if not thousands, of us. Most Fortune 500 companies have positions for people acting as liaisons between business and IT.

The International Institute for Business Analysis ( is creating a body of knowledge for the industry. The IIBA has published a description of the role and is opening local chapters in North America. —Barbara A. Carkenord VP of Training B2T Training L.L.C.


Bottom line on TopCoder
Regarding the article published in the January issue, I'd like to clarify a few misrepresentations regarding the data collected by TopCoder.

First, the article, “The world's best coders are …” (Page 31) states: 'Based on results from participation in the company's frequent algorithm and software writing competitions, the company currently ranks an American, with a score of 2768.80, as the best coder in the world.' This statement is incorrect; the score refers to an aggregate country rating, not an individual coder's.

The next sentence further implies that the additional countries listed in the last paragraph are individuals, rather than each being a representation of the country as a whole.

The paragraph concludes with the most puzzling statement of all: 'However, that national triumph for the U.S. is balanced by data on average performance among TopCoder contest participants. By that measure, programmers from the Republic of Korea are at the top of the top 10 ranking and the U.S. doesn't even place on the list!'

It is true that TopCoder does record and publish statistics that measure the performance of both individuals, as well as the countries they represent. However, the data that does, in fact, represent 'average performance among TopCoder contest participants' places the U.S. at the top of the list, rather than below the 10th spot as the article implies.

In addition, I am at a loss to explain the inclusion of the Republic of Korea, since its aggregate performance is not among the top 10 represented on TopCoder's Web site.

TopCoder specializes in running various types of programming competitions with one of the goals being to utilize the statistical data that is captured to measure relative merit of individuals and groups of individuals. Our rating system is at the core of our business and we certainly take issue when that data is significantly misrepresented. —Mike Lydon Chief Technology Officer TopCoder Inc.