Pejibaye Blog

Part of “Conquering Entropy” First, I have to admit some of my recommendations on this topic are based on my own experience and anecdata, and not so much backed by numbers, please take them with a grain of salt. Risk Management Shipping code is risky. Any change could break the system. Over the years, I have learned to get comfortable with managing risk. For each change that I ship, I do a little risk assessment beforehand.

Continue reading 

Part of “Conquering Entropy” Most of the issues I have with AI-generated code are related to trust. Do I trust the person who wrote this ticket? Do I trust that the engineer who opened this PR understood the ticket and guided the coding agent to implement it properly? Do I trust the coding agent’s implementation? Do I trust our test suite to catch regressions before they hit production? Do I trust our CI/CD to properly build, test, and deploy our change?

Continue reading 

A collection of ideas on how to manage entropy in codebases inundated with AI-generated code and survive in the process. What’s Been Changing Ever since coding agents became popular, my company’s monorepo line count has exploded. The number of pull requests (PR) has roughly doubled (conservatively), and so has the review burden on all engineers. People think the biggest problem now is that there’s too much code to review. But there’s way worse stuff!

Continue reading 

LocalStack is a great product. It’s basically an emulator for the AWS cloud that you can run locally. We rely on it for development and automated testing. However, nothing comes for free. Understandably, the company behind this amazing piece of software has some bills to pay. They changed the license to require a subscription to run their provided Docker images. If the license is not provided as an environment variable, the image simply fails to start.

Continue reading 

I think a codebase can tell you a lot about the people that work on it. From this specific example in particular, I felt very reassured that the team takes work-life balance seriously and never works on weekends. For context, the software interfaces with old-school banking systems that famously do not run certain processes on holidays or weekends. We had some logic to to an early return to avoid running a process on bank holidays, imagine something like the following:

Continue reading 