Teaching basic lab skills
for research computing

Inserting Software Carpentry Graduates into Coding Communities

One of the issues I'd like to see Software Carpentry tackle is what happens to learners after they've attended a workshop. Thinking about my own experience, I attended a workshop in Melbourne in February 2013 and then 2.5 years later published my first ever reproducible paper. (As I continually have to explain to impatient managers at Research Platform Services, yes it will take people at least that long to fully work these things into their daily workflow.) The difference between me and a Software Carpentry attendee who never goes on to adopt the best practices we teach is probably not that I read better books, attend better follow-up classes or find better online tutorials, but rather that I'm part of an active community. In fact, besides that initial Software Carpentry workshop, I can't recall a single book, class or tutorial that was hugely influential in my development as a computational scientist. That two-day workshop in 2013 introduced me to the core concepts I needed to know, and then I stuck at working to incorporate them into my working life because the people "around" me, in both the Software Carpentry instructor community and the Python in the Atmospheric and Ocean Sciences (PyAOS) community, were also doing it.

My thought for tackling this issue is therefore to insert Software Carpentry graduates into communities that are programming language and research discipline specific. In other words, plug them into an environment where doing and discussing computational best practice is the norm, and where these discussions happen in a language that is most familiar and relevant to them. In some cases these communities already exist (e.g. we could ask PyAOS if it would like to be the official community for Software Carpentry graduates who use Python in the atmospheric and ocean sciences) and in other cases we could approach people in the Software Carpentry community to help establish them (e.g. Software Carpentry graduates who do bioinformatics using R). These official Software Carpentry alumni groups would have a mailing list for general chat and questions (that people are automatically signed up to after attending a workshop), a blog, a discipline and language specific capstone lesson that they maintain and they may even host workshops at relevant conferences (e.g. PyAOS runs a workshop at the American Meteorological Society Annual Meeting every year).

What do people think? Is this an initiative that the Steering Committee should look at implementing, and do you know of any groups like PyAOS that might be interested in being the destination community for Software Carpentry graduates who work in their field/language?

Dialogue & Discussion

You can review our commenting policy here.