Teaching basic lab skills
for research computing

Software Carpentry and Data Carpentry on Podcast.__init__

A couple of weeks ago I was interviewed by the Podcast.__init__ team to talk about Software Carpentry and Data Carpentry. The request for an interview actually came in a couple of months ago and I asked Greg Wilson about doing it, telling him I was a big fan of the podcast. He said he'd be happy to have me do it myself, giving me the chance to star in one of my favorite shows. I decided to do it, in pursuit of my goal of promoting Software Carpentry and Data Carpentry (and maybe more public exposure for myself).

I'm so glad I got a chance to talk about Software Carpentry and Data Carpentry. The hosts were great to work with. In preparing for the interview, I realized what a unique space these programs occupy. The basics of software programming and data management are not taught by experts but rather by peers. The curriculum is also built and continues to grow by the efforts of these same instructors, based on their experience as researchers and scientists. I'm hoping the podcast audience will find it equally exciting.

Something that came up that I wish I emphasized more was the importance of "slowing down." We talked about the importance of precoding work—writing in pseudocode, drawing flowcharts, etc., before diving into the code. Whether it's getting scientists to specifically do any precoding work, or the larger idea of learning software skills at all, taking the time to stop and learn to be more efficient and productive has big long term payoffs.

One big thing I realized after taping—I don't have a "radio voice." I say "um" a LOT. It's not really a big deal when speaking live, because I can integrate those pauses using facial expressions or body language, or the old trick of pausing for a sip of water or to pick up a pen. I also feel like not using them at all makes my speech sound more rehearsed and less natural. When I'm a disembodied voice though, those vocal fillers become much more pronounced. I need to find a balance here.

A side note to close on—I think xkcd comes up in pretty much every conversation that's related to science and technology. In this case, it was in the case of "sudo" trying to tell someone to do something. I was talking about why I often prefer working with computers to working with humans. Once you learn their language, computers always do only and exactly what you tell them to do. When was the last time a human did that? If you don't know this web comic and you're a science and technology nerd—check it out.

Dialogue & Discussion

You can review our commenting policy here.