In defence of boring technology
Every project I have ever regretted started with the sentence “this is a good excuse to try X.” Every project that quietly shipped and kept working was built on things I already knew were dull.
The tax nobody budgets for
New technology charges you twice. Once when you learn it, and again every time it fails in a way you haven’t seen before. Boring technology has already failed in every possible way, publicly, and someone has written up the fix.
// The most underrated architecture decision:
const stack = pickTheThing(youAlreadyKnow);
The Postgres box from 2014: still running
Where the novelty budget goes
The trick isn’t avoiding new things entirely. It’s admitting you get maybe one novelty token per project. Spend it on the part that is actually your product, not on the database, the queue, or the deployment pipeline.
The parts of the system nobody talks about at dinner parties are the parts that let you sleep.
No account, no tracking. One vote per reader.