Holistic programming is a way of building software with bigger picture in sight without concerning with the syntax of framework or data structure. This sounds like antithesis of popular software engineering approaches lile OOP(object oriented programming), DP(dynamic programming) or FP(functional programming), etc.

Complexity & Human Limitations:

Humans have an inherent limitation based on their IQ(intelligence quest) to manage complex programming needs. The complexity of projects requires one to remember 100s of elements to make an error free and bug free system. But neither all projects have same complexity nor all programmers have same capability. Therefore a group of individuals bunched in a team have limitations to make a error free and bug free system. That is why almost every system has bugs and glitches, except the space shuttle OS¹ .

Project complexity benchmarks

Below is a summary of complexity in small, medium and large projects.

A small project has complexity of 5K, where as medium project has combinatorial complexity of 50 million. A large scale project has complexity of 500 billion or more.

When the project is split in parts, the complexity decreases up to 20%. And when the project is split to distributed teams, the complexity reduces up to 2%. Meanwhile the complexity of holistic programming stays at the scale of 10K.

Explore project stories of varied complexity at https://fastbuilder.ai/products.

Holistic programming is way to look at the bigger picture and build the whole system in one go like a iron cast.

How to do holistic programming

1. Firstly build a product story
2. Then build the coding prompt sets
3. Finally finetune the downloaded codebase

Step 1: How to write a product story

Step 2: Building the coding prompt sets

Sample promptset for Angular: https://github.com/FastBuilderAI/angular

Sample promptset for Reactjs: https://github.com/FastBuilderAI/ReactJS

Sample promptset for NodeJS: https://github.com/FastBuilderAI/NodeJS

Step 3: Finetune & Repeat

The generated codebase may have missing imports and other such issues. You should test and improve the codebase. You can also improve the AI provider or promptsets to improve the generated codebase. And finally the real benefit of holistic programming is to keep improving the fidelity of larger picture. By improving the product story, the product keeps improving at he same pace.