Louis building heuristics

when you want to build something, there are thousands way you can do so

depending on your goals, having fun, making money, reaching some milestone, building a company, here we will focus on solving a customer problem while building a startup

open source

1. build the MVP (e.g. something useful in the least amount of time)
2. devs can use it in less than 30 seconds from first attention
3. readme has between 100 and 400 characters
4. readme use inverted pyramid principle (e.g. most important info first) - people have short attention span

algorithmic

sometimes you may face some algorithmic challenge, like building something fast, that gives a good UX and the code does not become too ugly eventually

1. solve the problem in your head or recommended, written. Writing is the most powerful way to think. solve the problem like if you would do it manually as a human in the world
2. write the pseudo code in parallel
3. see in your head / written, with different inputs if it reach the appropriate output
4. build the actual code if it works
5. optimise

tricks:

• write everything in a single file at first, even if it’s 5000 lines - nowadays LLMs can easily split your working code in multiple files easily in < 10 seconds
• use unit tests
• use debugging mode and tinker with different values - follow the algorithm step by step and see if it aligns w ur human algorithm