Towards a software engineering qualification

NZ is taking part in an international effort to develop a software engineering body of knowledge, to define what needs to be known by those professionally engaged in writing software. One Auckland consultant thinks such an undertaking makes sense, but another disagrees.

New Zealand is taking part in an international effort to develop a software engineering body of knowledge (SWEBOK), to define what needs to be known by those professionally engaged in writing software. But there’s debate as to whether such an undertaking makes sense. Auckland consultant Robert Barnes thinks it does, but Ian Mitchell, also a consultant in Auckland, disagrees.

It matters

By Robert Barnes

The SWEBOK project, led by the University of Maryland, is trying to agree and classify a body of knowledge needed by software engineers — those who are professionally engaged in the business of creating software systems. It’s an international project, each chapter having been contributed by different authors. Chapter 8, on software engineering management, has been written by New Zealanders Stephen MacDonell and Andrew Gray, then at the University of Otago.

Like the PMBOK for project management, SWEBOK is not a textbook you study: it is a catalogue of subjects with which a competent software engineer should be familiar. It becomes valuable as a guide for course curricula, and a grounding for qualifications based on it. Already some colleges are offering courses based on SWEBOK, but there is still a way to go before it achieves the equivalent status to PMP, which is becoming an essential qualification for a project manager.

Why does the SWEBOK matter? The widely publicised failure of projects such as INCIS show that we, as a profession, still have much to learn. Engineers do not build bridges that fall over, largely because there are accepted practices that they must follow to prevent such failure. It can be argued that, if information technology is to be a true profession, like law, medicine, etc, then an accepted qualification based on an agreed body of knowledge is necessary. Without it, software engineering is just an art, with some artisans better at it than others but no real understanding of why they are, what is needed to improve, or even what constitutes “good” or “bad” development practice.

While PMP does not make you a good project manager, by certifying that you have passed an exam covering the subjects important to project management it does at least set a baseline. An equivalent of PMP that is rigorous enough to be worthwhile yet general enough (and at a high-enough level to outlast the rapid shifts in technology), would be useful to the IT market, provided that it does not lock us into approaches that are already obsolete.

While it will take time for a “software management professional” qualification to emerge, I would be very surprised if it does not. Whether this will be important or irrelevant will be determined by the market. If PMBOK and PMP is any guide, SWEBOK is going to be significant.

Barnes is a director of iE3 Group and national councillor for the NZ Computer Society.

The hidden agenda

By Ian Mitchell

Any documentation of knowledge must be good. So how can I espouse the opposite view? By saying that it’s not the body of knowledge I write against but the hidden agenda behind the term SWEBOK.

SWEBOK has been put together by a loose consortium of US universities. It is implied that professionals qualified in some sense in this BOK will be more effective in their work and more employable. But it also implies that there will be a qualification, a certificate students will pay for, a product that can generate profit.

Today there is a tendency for students to pay for qualifications for the value they add to career prospects — not for the joy of learning, the intellectual challenge and excitement or the stimulus that understanding solutions to problems can bring, but for the money. They do not care about the content.

A BOK covering computing will be a major undertaking and change with time because of the scale of the task and because examinations require change in course content. Therefore updating the BOK will be expensive, so the qualification will be expensive.

The course delivery method (lecturers) must provide the new content in the year leading up to the examination, so the cycle will be at least three years.

Today students often say “Tell me how not why — I don’t have time for that”. And some employers denigrate academic education. But problems of the complexity we handle in IT are not solved by following the book but by innovation, analysis, construction and knowledge, some of which is at the tacit level, not the explicit level.

Much more significant is the need for practitioners to be registered or licensed. In Texas software engineers must be qualified in the same way as any other engineer. IPENZ, the professional engineers institute, is reviewing this option now.

Then the cost of professional registration is added to the cost of qualification — and the fees to the end user increase accordingly. Did I mention insurance?

Pressure to learn the explicit documented knowledge and not to learn the tacit knowledge held by experienced practitioners can result in failure. I recently took part in an arbitration where each side had well qualified project managers but the project management failed.

No, you are not going to learn it all from a book or a BOK or a CD. And what about attending NZCS sessions where your peers show you what really works?

Mitchell is past-president of the Auckland branch of the NZ Computer Society.

Join the newsletter!


Sign up to gain exclusive access to email subscriptions, event invitations, competitions, giveaways, and much more.

Membership is free, and your security and privacy remain protected. View our privacy policy before signing up.

Error: Please check your email address.

Tags engineering qualification

More about Andrew Corporation (Australia)

Show Comments