Software Engineering Folklore Survey

I love this. Two university researchers are asking real world coders how they code. They want to learn whether the theory taught in software design courses is actually used in the real world. They’ve built a short (20-question) survey that takes less than ten minutes to complete, and will open source the data once it’s all in. I’ve sent in my answers and I hope all the programmers reading this do so too. One of the researchers was my lecturer for Algorithms, which must have been punishment for his sins in a past life. The other influential prof from my degree was one of the authors of Notes on Postmodern Programming, and I hope my work mixes the unexpected and the pragmatic as well as they have.

tags: ,
  • http://theotherthomasotter.wordpress.com Thomas Otter

    Nat,
    This may seem a bit cheeky but while you are in survey mode, I’m also doing a survey of software developers as part of my PhD in IT law. I’m looking at their knowledge and attitude towards legal issues relating to software.

    I’ve got over 500 responses, but I would love a final boost before I sit down to crunch the numbers.

    Details and a link to the survey here.

    http://itlawresearch.wordpress.com/

  • http://blogs.gnome.org/bolsh Dave Neary

    Aren’t they worried about the “wishful thinking” effect? Most engineers familiar of XP and SCRUM would like to write unit tests before writing new code, but don’t do it most of the time anyway. And while we might like to refactor classes if they get deeper than 3 or 4 levels, most of the time, we don’t ever get around to it.

    Are they hoping that people will be extremely self-aware, or are they testing the things engineers wish they did?

    Dave.

  • http://holloway.co.nz/ Matthew Holloway

    Done. I haven’t thought that much about class hierarchy in ages.

  • http://www.radar.com/nat gnat

    @Dave: they’re asking people for what they do, not what they wish they did. Like all surveys, they’re hoping for self-awareness. I guess the next step for them would be to analyze some open source code to see what class hierarchies look like in the real world, although that really just tells you what the committers/BDFL’s preference is.