Sign In

Communications of the ACM

Communications of the ACM


In the April Viewpoint ("Taking the Lead in Licensing Software Engineers," p. 27), Donald Bagert trots out the old chestnut about the need to certify software engineers.

Licensing and certification are such dimwitted ideas that if Texas (where Bagert is from) were to remain the only state to require them, the software industry in Texas would be a basket case in a few short years. I suggest this is the reason Bagert makes his case so urgently. The only way certification can not do terrible damage to the Texas economy is if the rest of the country can be convinced to require certification as well. The plea is reminiscent of the protests of communal farm workers in the Ukraine in the 1960s when they took to the streets to protest that certain farms were not under central government regulation, and thus operated at an enormously unfair advantage. The software industry over the past 35 years has transformed the world, bringing a change that is as important and as beneficial as the Renaissance. It has done so entirely without the help of regulators, certifiers, and licensers. The notion that things would now be improved by a huge dose of government-enforced regulation is not just wrong but preposterous.

The issue here is not certification, as Bagert would have us believe. The issue is decertification. The scheme proposed would allow governments and quasi-governmental agencies to decide that certain software engineers should be prohibited from working in our field. Some of the people that would be decertified are trusted and treasured employees of successful companies—employees whom the market has resoundingly declared to be worth every penny they are paid. These people would be decertified because they lack some aspect of government-dictated skills. Nancy Mead of the S.E.I. has suggested that one of the bases for certification should be an undergraduate degree. Drummed out of the corps would thus be a whole generation of young Bill Gateses who entered the work force (without benefit of college degrees). Companies like Yahoo have hired these youngsters, invested in them, and allowed them to help create an industry that is turning the world on its ear. But with certification they'll have to go. Sorry about that Yahoo.

I find it morally and ethically repugnant that anyone should prohibit anyone else from seeking to sell his or her services to those who would willingly buy them. I am offended that some of the senior members of our community should be willing to set themselves up as a guru class with such powers over their colleagues.

And all for what? To protect the poor old market from unscrupulous, unethical, and incompetent programmers. Right. Like the state bar associations protect us from unscrupulous, unethical, and incompetent lawyers. I say we let the market protect itself. It's doing just fine so far.

I propose we let Texas remain the only certification state as a controlled experiment in whether certification is good or bad. At the end of a decade or so, if there are any viable software companies left in Texas, let them tell us what the real effect of certification has been.

Tom DeMarco
Camden, ME

Don Bagert Responds:
DeMarco is attacking me for something I did not write. Near the end of my "Viewpoint," I stated, "Note that virtually nothing has been said about the benefits and drawbacks of licensing itself (there are many of each) ..." My point was also summarized further: "The licensing of software engineers has already started, and will likely spread to most of the rest of the U.S. ... Licensing will directly or indirectly affect the development of all new software professionals in the years to come ... The computing community should continue taking the lead in both licensing and accreditation, and everyone needs to get involved." That is, since licensing seems inevitable, let's make sure it is something that the computing profession can live with.

DeMarco has some very interesting views, including the fact that statements reflect his opposition to licensing in any profession. Although these remarks do not relate to my "Viewpoint," I would like to briefly clarify something concerning the use of the term "software engineer." Until September 1998, it had been illegal in Texas to call yourself a "software engineer" since the invention of the term "software." Now there is a process in Texas by which one can legally refer to themselves as a software engineer, while (as I understand it) there is still no way to do so in 49 of the 50 states. I'm not sure how that hurts the Texas economy.

Back to Top

Rule as Myth

Hopefully, programmers won't start implementing a 4,000-year rule to avoid the "Y4K bug" alluded to by Hal Berghel in "Digital Village," (May 1999, p. 3), because the existence of such a rule is itself a myth.

The rules for leap year in the Gregorian calendar as set down in the original pronouncement by Pope Gregory included only rules for the four-year, 100-year, and 400-year exception cycles. It was this definition of the Gregorian calendar that was formally adopted by English Parliament in the 1700s and implicitly by England's American colonies at the same time. There has been no formal adoption in either England or the U.S. of any subsequent modification to the Gregorian calendar to introduce a 4,000-year exception cycle.

Some later editions of Encyclopedia Britannica erroneously imply the existence of a 4000-year rule without giving any authority for the claim of such a rule. When I brought this to the attention of Britannica's editors in 1996, I received the following response:

"The Calendar FAQ to which you referred in your message gives credit to Sir William Herschel for the '4,000-year rule' for determining leap years; this rule, however, should be properly credited to his son, Sir John Herschel, whose comprehensive Treatise on Astronomy first appeared in 1830 [more current versions of the Calendar FAQ now include this correction. See].

"The Gregorian year is still approximately 0.0003 of a day (about 26 seconds) longer than the average tropical year. This difference of 26 seconds a year would amount to a day in 3,323 years. To correct this error, Herschel suggested dropping a leap year every 4,000 years. As you pointed out, other 'fixes' have indeed been suggested (for example, by statistician M.B. Cotsworth in 1914). The last sentence in the entry 'leap year' describes but one of the possible resolutions, and a recommendation has been made to reword this statement in the future in order to prevent any further confusion." — Sherman J. Hollar, Editorial Division.

Note that Britannica's research confirms the 4,000-year was a suggestion, not an adopted rule, and that the nearest multiple of 400 to the actual correction cycle would be 3,200 years, not 4,000. Serendipitously, since the Gregorian Calendar was adopted and in sync with nature circa 1600, the point at which the error accumulates to one-half day will hit around 3200, which would work nicely for implementing a 3,200-year cycle rule starting in 3200. We have 1,200 years in which to adopt appropriate leap-year modifications. Hopefully we won't wait until 3195.

Joel C. Ewing
Fort Smith, AR

Hal Berghel Responds:
It would appear that I am one of those people who place too much faith in the Brittanica. Mea Culpa

Back to Top

Earning Trust

The article, "Building Consumer Trust Online" (Apr. 1999, p. 80) is more concerned about trust persuasions. I don't believe the article mentioned that sellers need to earn the trust of buyers. Confidential information and sharing of information is simply not a part of this discussion at all. We already have an intermediate financial trust with credit card arrangements.

In all the discussions, the sharing of other information is always from consumer to seller, never the opposite. Who wants to trust such an arrangement?

Sellers almost never use the "shared" information to deliver better products or services. They almost always use the information to try to sell more products, generally of inferior quality, or sell service that is not delivered or for which they can't be held accountable.

Lack of face-to-face relationships in a seller-to-buyer situation generally leads to disappointing results. When I buy locally, the seller resides locally and has great concern about delivering, since the local attorney and police department are close by for me to contact. When I buy on the Internet, the seller is always in another state and has little reason to worry about reprisals if I get a bum product. As a result, the satisfaction level is much lower (both ways). And this all has nothing to do with shared demographic information.

By the way, I can tell you how most Web sites handle demographic data they collect: they sell it to the suckers who think it's valuable. The only demographic that counts is how much of your money they get. If they get enough, they come back for more (similar to the gambling palaces that provide free food, free drink, free lodging, and free limo service for the high rollers). After making my first purchase (telephone) from, I get a catalogue once every three weeks. Never once do they mention better products, better services, or better quality. They just want more of my money. The same happened after I made my only online purchase via TheZones Web site. I now get weekly specials emailed to me (spam). None of these specials offer better products, better services, or better quality (based on any kind of demographics). They made a sale and want more of my money.

We don't need anonymous buyers, just an honest business transaction that can be reconciled when it is not satisfactory. That reconciliation is already in place with existing credit card customer satisfaction services; lost, stolen, or failed, the credit card company resolves the issue for you since it holds your payment money. Demographics are not necessary. Demographics are valuable only to the seller to find buyers who part with their money the quickest.

Moreover, there should never be such a thing as opt-out. It should be illegal to force an individual to request opt-out. The only legal arrangement should be opt-in, by written signature only. If nobody opts-in, that is what sellers live with. Too bad for the sellers, but this is the only acceptable arrangement.

Finally, the statement: "A more consumer-oriented information privacy model will lead to commercially valuable relationship exchanges with important benefits for consumers and companies doing business on the Internet" is hogwash. This is strictly a marketing statement, biased in favor of the seller. The only information exchange is from the buyer to the seller, and the seller uses that to determine how to sell to you again. I have not seen any benefits provided by the seller to the buyer.

Ken Kashmarek
Eldridge, IA

Back to Top

Misleading Numbers

The News Track item ("50 Million and Counting," May 1999, p. 9) is very misleading. I recognize this is a citation of Yahoo, not Communications' own data. But clearly there were not 50 million Internet users in or around 1986. I believe from other reading that Yahoo's definition of "introduction of the Internet" can more accurately be described as "introduction of HTTP" or "introduction of graphical Web browsers" in the early 1990s. So using a more accurate definition of "Internet" as the cutover to TCP/IP, then it was more like 15 years between the introduction of the Internet and 50 million users. If one recognizes all the widely used predecessor networks (Arpanet, Usenet, Bitnet), it was more than 20 years from widespread introduction to 50 million users. Thus network adoption has not proceeded at a significantly different rate from other new technologies. In comparison, PCs had a faster adoption rate, even defining the Apple II as the first consumer PC. And the FCC's prediction that most everyone will have HDTV only eight years after its introduction is very dubious.

Hank Walker
College Station, TX

Back to Top


I found the articles in the May 1999 Communications very interesting. They have relevance to almost everything we do here at American Honda.

In the article, "Understanding The Seductive Experience" (p. 45), there is a misspelling of the name of the artist that created the "juicer." The artist's name is Philippe Starck. His items can be found at starck.htm.

Don Unger
Torrance, CA

Back to Top


Please address all Forum correspondence to the Editor, Communications, 1515 Broadway, New York, NY 10036; email:

©1999 ACM  0002-0782/99/0700  $5.00

Permission to make digital or hard copies of all or part of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial advantage and that copies bear this notice and the full citation on the first page. To copy otherwise, to republish, to post on servers or to redistribute to lists, requires prior specific permission and/or a fee.

The Digital Library is published by the Association for Computing Machinery. Copyright © 1999 ACM, Inc.


No entries found