Skip to main content

Solving problems

Knowledge is knowing where the answers might lie. Solving problems is knowing how to apply the knowledge you have to solve a thing.

Elon Musk has a great 5-step engineering process to solve technical problems.

I like to be methodical when approaching and solving problems I have.

For me solving a problem usually involves going through these four steps:

  1. Find what the problem I want to solve is.
  2. Break the problem down to smaller parts I can solve.
  3. Work on solving the task and nothing else.
  4. Solve the problem.

1. Find what the problem is​

Most often the difficulty of solving some task is simply due to it not being clearly expressed for me. I get the majority the tasks I work on from my task manager 2Do, GitHub issues and TODO:'s in the code. And I make sure that I understand fully well what the problem actually is before solving it.

2. Break the problem down to smaller parts​

Once I found and understand what it is I want to solve. I look into it more and see if I can try to break down the problem into smaller parts I can work on. Often times this might not be necessary and I can start working on the task immediately but in times when the task I need to solve is either hard or stated too vaguely, I break it down further into parts that I can pick up and work on individually.

Usually I break things down using MindNode. I break down a problem into series of steps I can approach solving and then go through each of the steps to solve what I originally wanted to solve, modifying and adjusting things in light of newly found insights.

It often looks like this. Where LA and epictools are two tasks/projects with subtasks.

MindNode lets you focus in on one of the nodes for better focus.

Can also use arrows with descriptions for more high level views. Above is example of one brainstorming session.

Often times I do the decoupling of the task inside the app I am working in itself, be it a code editor (as comments) or 2Do (as note).

Other times I sketch problems out in FigJam or Excalidraw collaboratively with others.

Sometimes I doodle things on iPad with Procreate or Notes apps.

3. Work on solving the task and nothing else​

I make sure to do the one task I have at hand with no distractions.

4. Solve the problem​

As mentioned in research, I mostly use Google for searching mixed in with DuckDuckGo. I heavily use Dash for documentation and Dictionary for searching the wiki. I also use IRC and various forums specific to the problem I am solving.

I often remind myself of how to effectively ask questions so as to respect other people's time and maximize chances of getting a good answer.

Sometimes I already know the answer given my past experience without conducting help outside and the only thing left is to apply the solution I have.

In some cases, a problem may be complex enough that it needs time thinking more about it. Researching various possible solutions and things you could use. And playing with possible solutions and ideas in your head until you reach something worthwhile you can spend time implementing. This talk by Rich Hickey covers this point quite well.

Notes​