Teaching basic lab skills
for research computing

Better Across the Pond?

I'm sitting in a packed room helping out at the UCL bootcamp, the first of a series to be run in the United Kingdom (Newcastle is next, and there are plans to run additional workshops in Oxford, RAL, Glasgow, Edinburgh and Bristol). I'm thinking about two things: why would you design a 60 seater lecture theatre with only one window for ventilation and why is Software Carpentry so popular in the UK?

Here are the statistics: the waiting list for the UCL bootcamp is 3x bigger than the capacity of the room; the (larger) Newcastle workshop is oversubscribed twice over; our main issue is making sure that tutors at bootcamps aren't double-booked.

So why is Software Carpentry seeing such levels of interest?

One thing I do know is that it isn't just the Greg Wilson effect — Greg is only teaching at this workshop, the other bootcamps are being led by people from the Software Sustainability Institute, SoundSoftware and the local institutions. Is there another explanation? I'm going to suggest three possible factors and would like to hear if there's evidence to support or contradict it, particularly if there are similar factors in other countries.

  1. The BBC Micro / Sinclair Spectrum effect
  2. Doctoral Training Centres
  3. Climategate

The BBC Micro / Sinclair Spectrum effect

Thirty years ago, two computers changed the course of the UK IT industry: the BBC Micro and the Sinclair ZX Spectrum. These machines, and the associated TV programmes and magazines, encouraged a generation to experiment with programming. Fast forward to 2012, and not only does the generation which grew up with PEEKs and POKEs form the technical backbone of the video games industry and companies like ARM, but that generation is bemoaning the lack of something similar in schools that has led to successive generations not having the basic understanding of software programming.

Doctoral Training Centres

Doctoral Training Centres were initially funded by the UK research councils to increase the research capacity in interdisciplinary research activities such as the life sciences interface and complexity science that are difficult to locate within a traditional University's departmental organisation. Increasingly, students at the centres receive training on specialist transferable skills which are applicable to their area, and many centres are teaching software development. This means that those researchers undertaking PhDs within these centres have had additional training over the rest. There are over 50 DTCs across the country and across many disciplines.

Climategate

When a server at the Climate Research Unit was hacked and thousands of emails and computer files were spread across the internet, the impact was not just in the exposure of poor working practices. They showed to the world the struggles of a scientist called "Harry" as he attempted to wrestle with difficult data analysis code. Now, even though there is no evidence that the code was producing incorrect results, the fact that it was difficult to prove the validity caused a crisis. In the UK ClimateGate wasn't just confined to the scientific press: national newspapers such as the Guardian, Telegraph, and even the somewhat sensationalist Daily Mail picked up the story. Suddenly scientists were all over the news, and it wasn't pretty.

So my theory is that given a culture in the UK where we have a group of researchers who didn't get the same direct experience of programming at school that their research leaders did, are seeing the next generation of PhD students snapping at their heels, and who definitely don't want to be part of the next coding scandal have realised that it's good to go back to basics, to learn not just what or how to programme, but why we programme.

And that is the ethos of Software Carpentry.

OPINION · UNIVERSITY COLLEGE LONDON

Dialogue & Discussion

You can review our commenting policy here.