I recently got a massive side project (that I’d previously hinted about) merged in to the master branch. A good time to reflect on why I even bothered.
Finding a reason
It’s not like improving the appearance of error messages can be counted as part of my thesis. Time in grad school is limited. Why do I put effort into tidying this code base?
Improving usability is, far too often, nobody’s job. Scientists have a rather bad habit of making code that works for us, writing a paper, releasing a link into the wild, and then… moving on to the next project. Publish or perish, eh? Releasing an update that improves progress logging won’t get you anything new to be cited; releasing a new flimsy tool just might. But then it becomes difficult to stand on the shoulders of giants and all that jazz.
The external incentives to clean up academic code simply don’t really exist. So you need to find other reasons.
Practicalities
If I need to explain myself to someone, there are ways I do so. For example:
- I, and other people I work with, are constantly using these tools; improving their usability makes our jobs easier. The less that someone has to ask the developers to explain an error, the more time both of us have to do our work.
- Having a low-brainpower side project gives me something to “relax” with. If my actual work is stuck or waiting for whatever reason, I can turn to my side project and still get something done.
- Chipping away at the perception of our tool being user-unfriendly might mean more people use it, which might mean we get more citations down the line.
But are any of those my real reason? Nah. My real reason is, as the blog post’s title suggests…
Joy
I enjoy my side projects. Working on them brings me calm, a sense of being useful, constant bite-sized accomplishments. Fixing small issues is relaxing. There’s a reason I do a lot of volunteer copy-editing for an Internet community that I’m fond of, and the reason is very simple: I like copy-editing. Code cleanup is similar.
The best reward is, of course, when I see my side projects jump into action. My new unit tests have already prevented a few errors from being introduced. These are the fruits of my labor of love. My labor of joy.
But if you’re not the kind of person who finds joy in this stuff, then maybe you can just re-read the previous section.