Regarding Warren Keuffel's article, "Decline and Fall?" (Interface, Aug. 2005): I'm currently a master's student in Boston University's software engineering program, and wanted to respond to Larry Bernstein and Barry Boehm's comments about shifting the focus of academic programs toward design and problem solving rather than development. I couldn't agree more. On reentry into the program in 2004 after a two-year hiatus from academic life, I found a major shift in what was offered in our handbook compared to the actual classes. Things like compiler classes and advanced data structures were replaced by software project management and design analysis of software systems—and I found it quite refreshing.
It's easy to pick up a book and learn about new products and practices in software development, but to actually research and see case studies involving real-life software companies and learn how to bring a product to market is highly beneficial to most software students who haven't had exposure to code of more than a couple hundred lines. In this way, students really get the feel of what's involved in such large-scale projects and learn some valuable lessons.
Matthew Powers
MS Student
Computer Systems Engineering
Boston University
South Boston, Mass.
THE POSTMODERN PROBLEM
Regarding Mr. Keuffel's questions in "Decline and Fall?": I see a couple of reasons for our country's waning interest in science in general and computers in particular. First, computer programming just isn't that much fun anymore.No longer can the lone coder have any hope of fame. The efforts of the brilliant developer are diluted by the rest of the team, which now has as much chance of failing as any other team.
Furthermore, our understanding of project management has exceeded our ability to actually put the solutions into practice. I've had a number of managers who had never coded. Just try talking to a nontechnical manager about such things as probabilistic estimating, burn-down charts, broadband Delphi and the need for periodic refactoring of code: "I don't care how sophisticated your excuses are—just get me that software." Any developer can see that faster CPUs and fancier languages won't solve this problem because unrealistic expectations always outpace what is possible. (At least we can count on the emerging nations to suffer the same fate in a few decades.)
Science in general faces a worse problem: Our nation has been in the grip of philosophical postmodernism since the 1960s. (It takes at least a generation for such a thing to work its way through a society.) This system teaches a number of things, including a rejection of absolutes and truth, and cynicism caused by technology's failure to deliver on its promises from the modern era. Such concepts are opposed to the aims of science. Couple that with our culture's loss of a motivating vision for the future, and the picture is complete: Why go into science if the goals are uncertain, and in some cases, unreachable?
Personally, I still enjoy software development (and detest postmodernism). I intend to adapt and remain in the computer field. But, in general, no amount of research money or incentives to major in the sciences will cure these problems. They're rooted in human nature—and those roots run very deep.
Name withheld
Senior Software Engineer
Mission, Kans.
CRUCIAL COMPONENTS
Regarding "Decline and Fall?": I got my CS degree back at its peak when only CS people wrote programs. Now all disciplines have the ability to do so.
I've recently built a new computer for myself in a weekend (once all the parts arrived). Could I build a software system in weekend? Why not?
Have we been just writing and rewriting the same software for the past 20 years, yet can't seem to do it any quicker or cheaper? How many of us have jobs integrating the Web to legacy systems?
Computer science and programming have become boring to a certain extent. I agree with Barry Boehm that the focus will be on the systems of systems. Now, we're concerned with building the right thing (for the customer) and making sure that it gets built according to a timeline so the value can be returned. Programming is becoming coding, much like a typist transcribing handwritten notes. It really doesn't matter if J2EE is better or not than .NET because they're both archaic. In the time it takes to install one of these platforms, I should be able to build something useful.
I'll always love the C language but also understand the popularity of Visual Basic. But we need the next level—something akin to building a computer by ordering a bunch of components. That level will allow me to come up with an application idea and put it together in short order.
Todd A. Hartle
Development Manager, Engineering
Digital Insight Corporation
Atlanta, Ga.
Warren Keuffel responds:
People are waking up to the value of component software and realizing there's no need to reinvent the wheel. Read about the FBI's take in my next column.
Warren Keuffel's "Decline and Fall?" rang true for me. I spent nearly two decades as a systems engineer working on the design and testing of leading-edge systems for the Navy before taking my current position as IS manager with a local school system. When I read Bernstein's quote about the "emphasis on requirements engineering, problem solving, design and software project management," I almost fell out of my chair—I couldn't agree more.
My main task is working with the vendor for our Student Information System (SIS). I've stressed Bernstein's four points, as well as software QA and configuration management, but it's as if I'm speaking an alien tongue. I've also had my fair share of problems with my customers in trying to define project requirements. People have lost the ability to separate requirements from wants.
Reversing these trends will not be easy, and I'm not sure how to do so, but I do have the following observations.
- Somehow, we've reached a point at which anyone who's ever written a "Hello World" program or a macro is considered a seasoned programmer.
- Business has trended toward specialization to squeeze out every ounce of efficiency. One unforeseen complication is the inbreeding this has caused. When I was an engineer in the Navy, the engineer was the one who rose to management. Here at the school corporation, it's the teachers who become principals. While every field needs its professionals, they aren't necessarily the best qualified to manage; indeed, someone with external experience often provides fresh perspective.
- Problem solving and requirements definition skills shouldn't be limited to software types. The workforce's lack of problem-solving ability is appalling. Just because you're becoming an accountant or an artist doesn't mean that you shouldn't take courses in problem solving, logic and computer use.
- If you can't find a solution from a conventional source, perusing some unconventional sources is worth the effort. Exploring other avenues is a powerful tool that can make or break a project.
- We as a society need flexibility to be able to make and learn from a mistake rather than change the parameters of the exercise to deny that the debacle occurred. Acknowledging and examining an error is much better than denying it and hiding from it.
I doubt that any of this will get fixed in my career and will undoubtedly be left for my children's generation to bear, but thanks for letting me vent some frustration.
Jay Leslie
Information Systems Manager
Monroe County Community School Corporation
Bloomington, Ind.
Warren Keuffel's "Decline and Fall?" is like many other media stories about the declining interest in computer science in America.
It's no accident that fewer students are studying CS these days. With the rise of offshoring, it seems ludicrous to study a field in which your chances of being replaced by a foreign worker rise by the week.
Bill Gates and other industry leaders cry about the lack of U.S. workers interested in CS, but at the same time lead the charge to offshore.
Cry all you want, but until there's a change in behavior by these companies, all that talk is just that: talk and PR hype to give an illusion they're interested in their domestic employees.
Joe Springer
Software QA Manager
Alameda, Calif.
In 1967, I went to Minneapolis to teach computer maintenance for Control Data Institute. We had full classes and two shifts. In 1970, the classes weren't full, because job growth declined. And, at what is now called Stennis Space Center, there's a cable tunnel going toward the never-built test stand for the Nova Mars rocket.
Why? The Vietnam War stopped government spending on everything but the on-ground military. The current war will do the same thing. Couple this with the fact that children today don't know that they're expected to work to succeed.
Keuffel has started to say what must be said, and he has an audience. Please don't stop short. The legacy of our generation depends on someone with the financial backing and guts to say what must be said.
Bill Cosby has been trying to get the attention of black children, and people are unhappy with him for doing it. The ethnicity must be removed, and all American children must be dragged kicking and screaming toward cleaning the house and studying in preparation for careers.
Name withheld
Software Engineer
Pascagoula, Miss.
I just read Warren Keuffel's article about the decline and fall of U.S. science and technology. The answer is simple: If you want people to go into computer science and other scientific fields in the U.S., you need to hire them here. Nothing raises my ire more than hearing the likes of Bill Gates and others saying that the U.S. isn't producing enough engineers. Yet, rather than hire American engineers and thus make these fields desirable, which would increase enrollment, they'll hire thousands in India, China and other points where they can be had cheap.
Those heading to college aren't stupid. Why should they enter into a field that, upon graduating, will possibly leave them in debt because they most likely won't be able to find a job? The decline in numbers of those entering engineering and scientific disciplines is a direct consequence of the shortsighted policies of our government and the greed of our multinational corporations.
Build it and they will come.
B. Binole
Senior Software Engineer
Cincinnati, Ohio
Warren Keuffel responds:
Hopefully our society will wake up and recognize that in the long run, outsourcing will cost us much more than it saves.
The simple answer to "Decline and Fall?" is this: No empire has ever returned from the brink of wealth and arrogance. The biggest difference between past empires and the United States is that the wealthy can now simply move to greener pastures.
It's unfortunate that our leadership has determined that wealth for the few is better than health for the masses.
Ted Fahrer
Software Engineer
Tyco Healthcare/Mallinckrodt
Cincinnati, Ohio
Since we have no idea what the world will be like in 10 or 20 years, students should take courses in mathematics, science, literature and history. If they do that, later they can learn whatever new technologies show up.
Of course, fundamental computer-science building blocks such as linked lists, queues, sorting and hashing are actually interesting in themselves, so folks should study those subjects, also.
What they shouldn't do is spend too much time focusing on knowing every detail about Visual Basic 6.0. You're only young once, so sponge up stuff that has a good shelf life.
Name withheld
ON TO ACCOUNTING?
"Decline and Fall?" glossed over the main factors for the drop-off in CS applicants: The industry is simply not the same anymore. Given the choice, most incoming students would not choose to enter an industry that is becoming increasingly hostile to be in.
With outsourcing/offshoring, the short "expiration dates" of many CS technologies and a seemingly endless stream of layoffs, I'm surprised it's not worse.
I've been seeing articles recently that accounting (of all things) is a rising favorite among incoming students. The reason? Enron, Worldcom and others have made it fashionable.
Robert Zotti
Program Director, Online Learning/CPE
Stevens Institute of Technology
Hoboken, N.J.
In "Decline and Fall?" Warren Keuffel's closing question, "What will it take to reverse these trends?" has a simple and historically demonstrable answer, but it's likely not the one he wants.
History tells that no nation occupies the king-of-the-hill position forever. Each occupant convinced itself that it was the last and greatest superpower before eventually decaying into an inevitable mediocrity.
We've already begun this phase, and no amount of editorial alarmism can change this downward acceleration. Consider this as not an IT issue but one facing all of our society, and it becomes clear.
David Fisher
President
Fisher Aircraft Corp.
Rochester,N.Y.
I was interested in Warren Keuffel's "Patently Inferior" article (Interface, July 2005).
As a small company that sits in both the open source world and the proprietary world, we're frightened by the spectre of a patent troll taking us on. A single attack could knock us out of business—and I'm sure that's the case for most small companies.
Whilst Mr. Keuffel's article did at least give a glimmer of hope for open source developers, it doesn't offer much hope for small businesses involved in proprietary software. I doubt many people will shed too many tears for small-scale software developers, though it should be remembered that many very useful products could end up bankrupting the author(s) if the authors are pursued for patent infringement. The revenues earned from many software packages would never be sufficient to defend against a patent infringement suit, never mind the damages awarded if the case is lost.
It's a sad day when the next generation of software entrepreneurs is killed off by patents—even sadder when many of the patents are given on spurious grounds.
The only glimmer of hope is that the European Union has rejected software patents. Given the United States's massive lead in the software industry, won't software patents severely limit its ability to compete in the longer term? Isn't there a danger that two wholly separate software industries could develop—with the U.S. dominated by a few large players and the trolls whacking start-ups with patents whilst everywhere else a more start-up-friendly landscape emerges?
Jack Hughes
CTO
OPENXTRA Ltd.
West Yorkshire, England
I just finished reading the August 2005 issue and must comment on Kelly Walker's letter concerning the Print Screen button ("Now They Tell Us," Feedback), in response to Alexandra Weber Morales's June 2005 editorial, "The Whack Award." "Useless" utilities exist because they often do something better than the standard fare provided with the OS. The first time I saw a business analyst documenting a system by using the Print Screen-Switch to Word-Paste-Switch to Application method, I was horrified at the prospect of making 50 screenshots that way. It took me all of five minutes to find an acceptable and free utility that would take a screenshot and save it in a file. That was years ago, and to this day we can't estimate the savings in time and frustration that this and other utilities have provided. There are many useful utilities that save keystrokes, make a task easier and foster accuracy; one must just spend the time to find them and recommend their use.
Bruce Neild
Software Product Manager
St. Paul Travelers
Hartford, Conn.
On reading the praises of Stan Kelly-Bootle ("Cheers to the Industry's Boswell!" July 2005) and resampling the SKB sampler, I'm reminded just how deficient my SKB parser must be. In the universe where his writing makes sense—where subject, verb and object cohere and signify, I remain unsure of many things. It's a Heisenbergian uncertainty—either I sense where he's going, but I can't determine where he stands; or alternatively, I understand his position precisely, but I haven't a clue where he's headed. But what do I know anyway? The very first book I stopped reading was Joyce's Ulysses—for the very same reason.
Todd Tieger
Princeton, N.J.
Did I miss something in Mike Riley's article "May the Source Be withYou" (Special Guide to Configuration Management, Aug. 2005)? A discussion of source code control without mention of SourceSafe? I'll admit I didn't read every word of the article, but I didn't see it mentioned in the tables.
Name withheld
Mike Riley responds:
I indirectly referenced SourceSafe via my selective mention of SourceGear's Vault, a product that extends Microsoft's aging Visual SourceSafe to distributed team contribution standards with Vault's many Visual SourceSafe enhancements. Microsoft will likely redeem itself with the introduction of its forthcoming Visual Studio Team System, but for those seeking an enhanced and less expensive SourceSafe experience, SourceGear's Vault remains one of the few viable alternatives.
I really liked the August 2005 phone issue ("Reinventing the Smart Phone" by Rosalyn Lum and "Visions of VoIP" by John Ravella, Joe Falcone and Gareth Meyrick). I'm working on mobile devices right now and really enjoyed the articles and their perspective, as well as the interviews with the folks from different parts of the world. Great job!
Bernard Golden
Chief Executive Officer
Navica
San Carlos, Calif.
EXPRESS YOURSELF
Software Development welcomes readers' letters. Writers must provide their full name, title, company, mailing address and telephone number. Letters are edited for clarity and length. Although we publish only a portion of the letters we receive, we value the views of readers who send us their comments.
Send letters to FEEDBACK, Software Development magazine, 600 Harrison St., San Francisco, CA 94107. Send e-mail to [email protected] or [email protected].