Англофил (valchess) wrote,
Англофил
valchess

Цитатник Англофила по выходным - 10: статья об умирании Computer Science

Полторы недели назад посетил я University of Warwick. Это один из ведущих британских университетов: во всех league tables он занимает места в районе 5 - 7 и считается самым успешным из т.н "новых" университетов. Знаменит своей ориентацией на research и многогранными связями с промышленностью. А если говорить об инфраструктуре (о кампусе, в частности), то соперников в Британии у него почти нет (один Arts Centre чего стоит!). Когда-то (более 10-ти лет назад!) я провел там на кафедре Computer Science целый год - нося при этом гордое (можно сказать, королевское) звание the Royal Society Research Fellow. А ныне прибыл в качестве guest lecturer ознакомить тамошних студентов и аспирантов с некоторыми современными главами компьютерной графики...

Так вот: отчитав свое и водрузившись с чашкой кофе в кафедральной Staff Room (замечательно оборудованной, как и все их новое здание), я обнаружил, что находящиеся тут же коллеги с большой горячностью о чем-то дискутируют. Как выяснилось, они обсуждали статью в THES - Times Higher Education Supplement (это толстая еженедельная газета, посвященная британскому высшему образованию), которую я и хочу предложить в рамках данной рубрики, где размещаются тексты на языке оригинала.

Несколько слов в пояснение. Уже несколько лет в Альбионе наблюдается тенденция снижения желающих обучаться по специальности Computer Science - собственно, с 2001 года количество заявлений на прием в масштабах страны снизилось на 50%. Вот в этой полемической статье и утверждается, что такая ситуация возникла закономерно, и надо, наконец, осознать проблему. Традиционная Computer Science (бастионом которой и является кафедра в Уорвике, котирующаяся среди себе подобных очень высоко, но действительно испытывающая ныне проблемы с количеством принимаемых студентов) все больше становится оторванной от практики дисциплиной. Все эти высокоумные математические основы, формальные методы, исследования алгоритмов, научно обоснованные дисциплины разработки ПО ныне, дескать, все более теряют актуальность. IT индустрия все более превращается в сервисную отрасль, продукты которой уже не требуют для своей разработки знания тех дисциплин, что изучаются в рамках традиционных математизированных курсов. Ныне все собирается из готовых кирпичиков, и IT-специалистам надо в первую очередь понимать бизнес-логику, овладевать e-commerce, уметь работать с людЯми...

Собственно, в России - без больших дискуссий - процесс коренного качественного изменения подготовки компьютерных специалистов уже давно запущен. В явочном порядке. Всем известно, что российские студенты компьютерных специальностей уже с третьего - четвертого курсов работают full-time по специальности (в Альбионе такое практически не встречается). Соответственно, они не имеют возможности надлежащим образом изучать продвинутые специальные курсы (хотя, конечно, успешно их сдают). Зато еще до окончания вуза становятся востребованными специалистами, способными успешно работать. Правда, есть мнение, что и продукты, такими специалистами разрабатываемые, оставляют желать (о том, что российской Computer Science практически нет на мировой научной карте - наверное, в том числе и потому, что выпускники даже ведущих вузов не знают базисных основ своей науки - я и не говорю)...

Итак, привожу упомянутую статью (вызвавшую действительно большой резонанс, и не только в Британии), а также один из откликов на нее, опубликованный в следующем номере THES и подписаный полусотней профессоров, среди которых большинство руководителей кафедр различных университетов. В качестве бонуса будем иметь достаточно типичный пример газетной полемики.

ERASE OLD PROGRAMME AND LAUNCH NEW VERSION

by Neil McBride, principal lecturer in computing at De Montfort University.

Computer science is an out-of-touch, dying discipline. It must embrace interdisciplinarity to survive

Go into a computer science class today and you may well be entering a time warp. Warm, comforting, cocooned from the changing commercial world, you will hear the pure tones of a computational discipline. Algorithms, operating systems, data and program structures appear before you on the whiteboard. There will be an air of certainty and assurance. It's a scene that has remained unchanged for 30 years. And one wonders: have computer science academics lost touch with the real world?

There is no doubt that computing is in crisis. Applications for university courses fell 11 per cent in 2006 alone. And it is a global problem. The number of US students choosing computer science dropped 39 percent between 2000 and 2005. In the UK, computing departments are combining with business or downsizing. Their research base is weak, lagging behind that of industry. Concepts such as service-oriented computing appear on the university agenda only after becoming current practice in industry. It is a dying discipline, an old man who has run his race well. He changed the nature of human existence, but perhaps it is time to let him go.

The environment within which computing exists has changed. Computer science has lost its mystique. In the 1970s, it was the realm of the experts. Programming was difficult, laborious and required significant training. A high priesthood of mathematicians and physicists attended the mainframes in air-conditioned temples.

Today, it is part of the fabric of society. We are all involved with computers in business, leisure and the home. Programs that were part of university syllabus now run in primary schools. Web design, database building and application development have moved out of the laboratory into the street. The public perception of computing is that it is neither unusual nor inherently interesting.

In industry, the nature of information technology has changed. Business applications are rarely constructed from scratch. Companies do not hire armies of programmers. IT is seen as a service, computing a commodity. Skills in IT departments centre on procurement, integration, evaluation, contract management and customer interaction.

This focus on continuous services contrasts dramatically with the philosophy of most computing curricula, which concentrate on building systems based on the systems development life cycle, a linear process of determining feasibility, analysis, design, programming and testing. It is an approach that is of declining relevance to industry.

For computer departments in universities in Europe, the US and Australasia, globalisation constitutes an even greater problem. Since the software can be transmitted almost instantaneously, why develop it in expensive facilities in the West? The very paradigms and skills that are dear to British computer scientists have been taken over by countries such as India, where 100,000 skilled English-speaking computer scientists graduate every year. Companies such as Microsoft and Intel produce innovative software in India as a matter of course.

As enrolment plummets, the computer scientist blames the customers. If only potential students saw how exciting computing is, they would flood back. But schoolchildren are not stupid. They can read the writing on the wall. It is the computer science academic who is to blame for the state of computing. First, they have failed to engage with society, to properly acknowledge the social embedding of computing and its changing role. Second, they have not put sufficient effort into searching for new paradigms. Like physicists at the start of the 20th century, they are not thinking outside the box. Third, they have failed to engage with industry.

Look at the contrast with molecular biology, where academe leads industry. A biochemistry PhD is a key to an industrial post. In computing, a PhD is not a gateway to industry. It may even be frowned on. As an old oak tree falls to the ground, buffeted by storms, rotting inside, computing should make way for new saplings to emerge. The new discipline will be complex and divergent, based on relationships. Instead of being narrow and blinkered, it will be outward looking, drawing ideas from biology, design studies, history, geography. It will be creative and risk-taking, an interdiscipline rather than a discipline.

In 30 years, there may be few computer science departments. Most will have given way to interdisciplinary studies and services departments that produce graduates who can control, develop and manage the IT resources an organisation needs.

So let us say goodbye to the respected old man. Let us look to the new generation of interdisciplinary computing. Let us lead the young by the hand and create new paradigms and new directions for a new discipline.

DEATH OF COMPUTER SCIENCE GREATLY EXAGGERATED

by Jose Luiz Fiadeiro, Professor and Head of Department of Computer Science, University of Leicester

Neil McBride’s piece on “Why Computer Science is dying” (February 9) is stimulating but utterly wrong, both as to facts and its unsupported claims: there are more CS students in the UK now than in all the standard sciences put together, and few would argue that basic physical science is dying, or losing out to industrial research, just because it is sometimes hard to recruit undergraduates to it from our current school system. McBride’s complaints may best be seen as a description of what has happened in universities which do not do CS research (and that does not apply to his own, of course), but his disbelief that Microsoft hires CS PhDs suggests he himself knows little of that world, and Bill Gates’ recent speech (Toronto Globe and Mail, February 8) says exactly the opposite. There is an oddity in McBride rubbishing the products of CS departments but telling us, with apparent approbation, that Indian universities produce “100,000 skilled English-speaking” graduates every year, having taken over “the very paradigms and skills that are dear to British computer scientists”. If the Indians are right then so must we be.


There is no difficulty finding out the basic facts about all this: the EPSRC website shows that its International Review of UK Computer Science found university CS research standards excellent; the creators of web search engines like Google continue to come from University CS departments, and UK CS Departments did important work that made Skype and MSN Messenger possible. All market surveys show huge and continuing demand for well-trained CS graduates in the UK and US, a demand that is indeed being met by India, partly because of pessimistic beliefs like McBride’s in this country.


There are real issues that his piece touches on, issues well-known in the field and much debated: there is the gap between the CS core, taught in good departments, and much in demand in industry, versus much softer IT/ICT disciplines, which involve some business training, and much use of commercial packages but little rigorous programming. E-skills degrees of this sort are now being put on in many places, and one must wish them well: but it is not yet clear what the future of their graduates is, because they are neither one thing nor the other, and recruiters know this. But, and here is the crux, promotion of these new degrees does not require or support an attack on CS itself in the way McBride has chosen.


[FOR THE EDITOR: The 51 signatories are all Deans, Heads of Department or Professors of Computer Science at the University shown for them, and are evenly spread between “new” and “old” institutions—a point I would appreciate being made if the letter is printed but not with all signatories listed.].
Tags: Наука, Образование, Цитатник
Subscribe
  • Post a new comment

    Error

    Anonymous comments are disabled in this journal

    default userpic

    Your IP address will be recorded 

  • 27 comments