How To Learn Technical Things
I see developers, especially juniors, stress about learning new things. Often they compare themselves to a developer who already knows what they are learning and they feel like they come up short. I do this to myself, and I know how frustrating and demotivating it can be.
The people that know the concept you are learning went through the same process that you are going through - often multiple times - to get where they are today.
The actual mechanics of learning are the same for everyone: you poke around, you push the edges of what you know, you make mistakes, you do dumb things, you struggle to understand, you apply it to problems you are interested in, and knowledge grows over time.
Some developers seem to be better at learning new things than others. This feels loosely correlated with experience, but experience isn’t the only factor. What techniques does a skilled learner use? Here are a few ideas shaped by my own experience and observations.
Good learners make mistakes. Mistakes are key to exploring the boundaries of your knowledge and discovering where your assumptions are incorrect. Everyone makes mistakes when learning, but the best learners make those mistakes fast. They aren’t afraid of breaking things or doing things wrong - they relish it. By breaking things, they figure out how to fix them. By doing things wrong, they learn the right way.
One of my favorite examples of this is Julia Evans’ blog posts. She is amazing at diving deeply into things that are a bit intimidating to me, like the Linux kernel, the Java GC, or cool things you can do with strace. Her blog posts are often stories of her explorations with some new system concept, including all the things she tried that didn’t work or all the things that went wrong. The posts usually end up with a list of things that she learned, and I learn things from reading them.
Asking a question exposes your lack of knowledge. This can be intimidating, You are admitting that you don’t know something, which can feel a lot like exposing a weakness. Despite being scary, asking questions is a great way to expose fundamental misunderstandings or make you aware of unknown unknowns.
My friend Aimee Knight is a great example of this. She keeps a list of questions on a piece of paper. Every week she pulls someone aside to go through her list. She is great at learning.
Get rapid feedback
Good learners are relentless in the pursuit of feedback. This might mean unit tests if you are playing with a new technique or library, a REPL if you are learning a new language, or building a small demo if you are learning a UI framework. The common thread is that they try to set up a feedback loop that can very quickly answer the question “is this right?”
My friend Dallin Osmun is amazing at this. He wanted to learn GraphQL. My instinct is to spend hours reading docs and understanding concepts. Dallin focused on the installation instructions and immediately started building a GraphQL server and a client that consumed the data. He knew that he faster he could get some code running, the easier it would be to get feedback on his understanding. Dallin is a good learner.
Good learners build knowledge iteratively. They accept that they don’t know the perfect way to do something, but they still get things done with the knowledge they have now. After working in an environment where they are experts it can be uncomfortable to be a beginner again. They embrace and push through that discomfort.
Compare it to what you know
Good learners often take shortcuts by comparing to something they already know. This is one area where being experienced can help. If you have seen many different programming languages you discover that most languages are a mixture of several different concepts, most of which show up in other languages. If you already understand immutability then Clojure or Elm or Haskell, all languages that heavily feature immutability, get a bit easier to learn.
Beware. Sometimes these shortcuts can mislead you. You might dismiss something good because it bears a passing similarity to a technology you had a bad experience with previously. If a new tech reminds you of something you already know you might skip understanding it, only to have it bite you later on when you uncover fundamental differences.
Good learners keep trying. It often takes a few tries between with a technology to break through to understanding. Struggling to learn something doesn’t mean you can’t learn it. It means you are growing. It took me a few tries to learn React, node.js made no sense to me for a few months, Clojure was baffling the first two times I tried to learn it (how can you program without mutating data?), and I still haven’t quite cracked Haskell. Keep going.
If you want to hear more about this, Dave Smith and I discussed it in Episode 16 of the Soft Skills Engineering podcast.