The Devil is in the Details

I had never given much thought to the distinction between being a big idea person versus one who is detail oriented. Certainly I lean towards the big idea, but I can get very obsessive about small things such as cookie decorating or setting the table for dinner — apparently I’m only pay attention to the details of food or things relating to the consumption of food.

Thus far in my experience as a front-end engineer, this penchant for the broad scope has not created many problems for me. Certainly, my CSS is a little unpolished if a spec is not provided, but this trait has allowed me to visualize the structure of larger projects without getting bogged down by minutia.

I was recently tasked with contributing to an open source project as a group and was thrilled to finally get the opportunity to give back to the tech community in some way. This was also our first experience of diving into a large, unfamiliar codebase in an unfamiliar language.

The assigned repo offered a substantial ReadMe which we skimmed to get the gist of the product — a website for a large conference here in Denver. Then it was off to the issues queue to choose an issue that was the right size and complexity for junior developers. We chose a simple issue: change an input field to a dropdown with predefined options.

Navigating Rails was probably the easiest part of this entire assignment. We sifted through files with unfamiliar suffixes figuring out which ones were tests and which ones actually related to our chosen issue. Identifying patterns becomes second nature as a developer and we quickly found the proper syntax to change our input to a select menu. These are all big-idea tasks and actually addressing the issue at hand took all of an hour.

We had to carefully read the PR guidelines in detail — after all, this was our first time contributing to someone else’s work and we didn’t want to do it wrong. Surprisingly, many tests were failing in files we hadn’t even touched. Locally, the site was fully functional, so we rationalized that these failing tests were not our fault and submitted a PR.

Obviously our PR request was graciously declined with a request to investigate the failed tests. It was pretty ridiculous to think that the codebase would merge a PR that is failing tests and the CI build. I’ll chalk my optimism up to my lack of experience, but now I know better. After hours of trying to locate the reason the tests and the build were failing, we realized we had not followed the ReadMe very carefully. We were missing an entire file of environment variables.

I’m not going to make this post much longer, for fear that other ‘big idea’ people will lose interest, but in the process of contributing to an open source project, I realized that I’m going to need to get a lot more detail oriented when it comes to code. I just hope that I can balance it with my obsession with pretty food.

Related posts:

Getting stuck in a technical interview happens to the best of us. In the Google interview, that happens by design. A good technical interview will test how well you can apply what you know to…