The Small Things That Make a Difference

I’m trying to create a habit out of learning something from my work every week. I try to observe the work that I’ve done and try to see if there’s something I did that I could have done better.

One of the main things I’ve encountered recently is the topic of “smaller feedback loops”. For anyone that’s reading developer articles, this wouldn’t sound like a new concept. It’s a big factor many people are pushing, that you should try to create fast feedback loops to verify that what you’re doing is on track, and you should use this feedback loop often.

One thing I noticed for myself recently was a feature I was doing that required calling an external API. I did the mistake of implementing all of the “easy” steps first, namely the wiring in my application and other stuff that was well known to me, to then finally fire it up in the staging environment… and see my call to the external API fail.

What I realized there was that when you have these kind of external dependencies that operate as unknowns for you, it is worthwhile trying to call them up front to make sure you’ve understood how they work. Simply by using curl it would have been possible for me to discover the faulty assumptions I’d made about the API.

In the end it was no big drama. There were some small adjustments that I needed to do, and then the code worked. But it was a good lesson that when you start implementing something, make sure that you understand your integration points with the outside world first, because you might very well have made assumptions that are false that change everything.