Temporal Time-Skipping: The Clock You Didn't Know Existed
How Temporal's time-skipping test server works, why it broke my workflow tests, and the mental model that finally made it click.
10 minute read
Categories
30 pages
Technology
Cursor Is Brilliant—And Here's Why I Still Hate Using It
When you need control over fanciness: why the smartest AI IDE fights serious engineering work, and what actually matters for pragmatic developers.
12 minute read
The Two-Agent Method: Building Features Without Drowning in Context
A practical approach to using AI coding assistants that actually works for complex codebases - by understanding how context windows decay and structuring your workflow accordingly.
10 minute read
Where Stateless APIs Meet Stateful Databases
The discovery of the boundary between stateless apps and stateful systems
7 minute read
Why AI Cannot Engineer
We need to remember there is a difference between engineering and programming.
12 minute read
AI and the Zero-Sum Fallacy: Why Tech CEOs' AI Talk Might Mislead You
Tech CEOs claim AI augments humans, but their actions tell a different story. Learn why their zero-sum reality doesn't have to be yours—and how to build AI job security on your own terms.
12 minute read
What Actually Is ASGI? A Journey from Confusion to Clarity
Demystifying ASGI by following one developer's journey from framework confusion to understanding the bare-bones protocol that powers modern Python async web servers
13 minute read
Litestar DTOs: The 'Advanced Feature' I Didn't Need (And When You Might)
I thought I was doing it wrong by not using DTOs. Turns out, sometimes the simple approach is exactly right—here's how to know which to choose
9 minute read
From Pydantic to msgspec: A Migration Story About Validation and Control
What I learned moving from FastAPI to Litestar—and discovering that validation doesn't have to be magic to be powerful
6 minute read
Stop Letting AI Rewrite Your Entire Codebase (And Start Coding Again)
How I went from burning $400/month on AI chaos to actually shipping quality code with LLMs as my assistant, not my replacement
10 minute read