engineering
Every public API eventually has to evolve in ways that break clients. The three common strategies — URL path, custom header, and date-based — each solve part of the problem and create different long-term costs. The honest comparison after running production APIs through multiple breaking changes.
engineering
Tests that hit real third-party APIs are flaky and slow. Tests that mock them by hand drift from reality and miss real bugs. The patterns that work in between are fixture-based recording, contract testing against the vendor's spec, and a small set of disciplines that keep the mocks honest.
engineering
Most product teams pass through a phase where they build an internal SDK for their own API. The intentions are good: to encapsulate authentication, to provide typed bindings, to give developers a faster path to a working integration. The execution often goes wrong in predictable ways: the SDK lags the
engineering
API quotas are the load-bearing constraint of any SaaS pricing model. Get them wrong and either the free tier abuses your infrastructure or the paid tier feels arbitrarily limited. The honest framework separates rate limits from quotas from billing meters, and treats the metering layer ...
engineering
Webhook payload schemas drift over time, and every drift is a potential break in your customers' integrations. The discipline of versioning, the patterns that minimize churn, and the migration path that respects the asymmetric cost of breaking changes paid by the customer.
engineering
The webhook is the most fragile API surface most products ship. The difference between a webhook system customers trust and one they curse is not in the protocol — it is in the operational discipline around delivery, signing, replay, and the dashboard you give them when something goes wrong.
engineering
The choice between Server-Sent Events, WebSockets, and long polling is not really a choice between protocols. It is a choice about which complexity tax you want to pay, and where in your stack you want to pay it.
engineering
Every API hits the moment when a change is necessary, the change is breaking, and clients in the wild cannot be updated atomically. The wrong response is to deploy and pray. The right response is a versioning strategy chosen before the first breaking change, not after it.
engineering
Time zones are an infinite source of bugs because nobody believes they will be the one to make the obvious mistake. Here is the short list of decisions that prevents almost all of them.
engineering
Offset pagination is fine until your table is large or your data shifts. Cursor pagination is harder to introduce after the fact. Here is the small set of decisions that determines whether your pagination scales.
engineering
The rate limiter in your starter template will not survive a real traffic spike. Here is what actually works in production, and the small set of decisions that matter more than the algorithm.
engineering
You don't need Redis, RabbitMQ, or Sidekiq for most background work. A 60-line Python loop and a SQLite table will carry you further than you think.