Solve the task

• Often, when we understand a task, an idea of how to solve it pops into our minds and the way from there to going ahead and coding.
• However, there are many things one should take into consideration when deciding on a solution, so going with the first solution that you had in mind without critically examining it will very often get you to either a solution that is not optimal or a solution that just doesn't work.

Why spend time on planning a solution

• No reinventing the wheel.
  • If we won't be aware of what we already have in our project abd what is available to us, we might end up writting a solution that already exists and is available to use.
• Write compatible code
• Not waste time on implementing suboptimal solutions
• Not find later that the solution doesn't work

Solve

• Define technically
  • What are the scope?
    • What already exists / What are your task boundaries
    • What needs to be done
  • What are the context?
    • Parts of the system that interact with your code
    • Code you can learn from or use
      • Some functions similar to the one you're trying to write, which you can get inspiration from
      • Maybe someone has already done what you're doing but differently, and reading his code or talking him could help
    • Coding standard
• Find optimal solution

   - What are some relevant solutions?
   - Which solutions is the best?
       - Finding all the criteria
           - Cost (time + money)
           - Performance
           - Forward compability => Make future tasks easier
           - Simplicity
           - Security
           - Completeness => Handle all cases
           - Conventions
           - ETC
       - Compare all the solutions by the criteria

  We can also apply this to other people, when we discuss about a problem, we can help them to get to the correct solution by asking them these questions: What are the different ways to solve the problem, which criterias are relevant to this problem and which way does he/she think is the best based on these criterias?

• How can we be certain it will actually work for some solutions?
  • Use a POC, basically means you find the problematic part in your solution and do the minimum needed in order to make sure you know how to handle it
  • Might want to consider building a test for it and test with different input to see if the output matches what we expect

Documents

• You want to make sure you write down all you know about the tasks and solution just so that you won't forget it by the time you need it.
• [[Understand the task#How to remember what you understand]]
• Can also you pseudo code