An iterative development process

README.md

A project/algorithm/feature iteration method

1. Understand the problem
   • This step is often squishy but always important.
   • What is the goal of the project/algorithm/feature?
   • Note: bugs may be out of scope and require a different method. See bug hunting.
   • The definition of "done" starts here.
   • You will often bounce between Steps 1 and 2. Beware of starting Step 3 too early.
2. Plan it out
   • Is the project/algorithm/feature documented? Pseudocode, drawings, flowcharts, etc.
   • This is a great time to get creative. Break out some markdown and watercolours.
   • Include all stakeholders that need to be included. The act of documenting is just as important as the documentation itself.
3. Brute force a solution
   • Many coders get stuck here, especially under a deadline.
   • This is a great time for paired coding, code mobs, etc.
   • git commit when you've made progress that doesn't break anything.
   • Take a short break (within the day, preferably) before Step 4.
4. Walk-through/test the solution
   • Add To Do comments directly to the code in case you leave the project on the shelf.
     • Is anything broken?
     • What can be improved?
   • This is a great time for a code review if there's time/preparation.
   • Don't change code unless it's quicker than writing a To Do to fix it later. Stay focused on the walk-through.
   • Perfect code doesn't exist! The level of optimization is up to you (or your boss/client/team).
   • git commit when you're ready to work on the To Dos.
   • Optional: take a break before Step 5.
5. Optimize the solution and continue to the next project/algorithm/feature.
   • Prioritize the low hanging fruit (easiest improvements with the most impact) and work your way up.
   • Avoid the temptation to start a new project/algorithm/feature before you git commit your optimizations. Don't fall into any rabbit holes.
   • Take a break before continuing.

Attributions

• Writing CSS Algorithms