Randomness, and the bugs you cannot reproduce
The bug that happens once a week and never in staging. A systematic approach to the class of problem everyone handles badly.
page 2
older dispatches from the README archive.
What a staff engineer does all day
The role is genuinely ambiguous and that ambiguity is load-bearing. An attempt at a concrete description.
Licenses: a practical guide for people who ship
Not legal advice. A working engineer's map of what the common licenses actually require you to do.
Distributed tracing that people actually use
Most tracing deployments produce beautiful waterfalls nobody opens. The difference is three implementation details.
Refactoring under an agent
Large mechanical refactors are the clearest win available from coding agents. Here is the process that keeps them safe.
The World Cup is the largest load test ever run
A month of synchronized global demand across three countries, sixteen cities, and every streaming platform at once.
WWDC 2026 and the platform that keeps its own counsel
New OS versions, more on-device model surface, and a developer relationship that remains complicated.
Static analysis is finally worth the false positives
The tools got dramatically better while everyone was ignoring them because of a bad experience in 2015.
Time zones, and why your calendar code is wrong
An instant is not a time. A time is not a date. The rules change by government decree with a few weeks' notice.
Rewrites: when they actually work
The received wisdom is never rewrite. The received wisdom is mostly right and has three real exceptions.
What a great bug report contains
Six fields. Most reports have two. The difference is measured in days of engineering time.
Edge computing, honestly
Running code close to users is a real win for a narrow set of workloads and a complication for everything else.
Model evaluation for people who ship
Not research benchmarks. A practical harness you can build in a day that makes every future model decision an hour instead of a week.
Idempotency is the only distributed systems concept you need
Not the only one. But if you only internalize one, make it this one, because it makes most of the others survivable.
Platform teams that don't get resented
Internal platforms fail for predictable reasons. The successful ones share four properties.
Open source funding: five models, honestly compared
Donations, foundations, dual licensing, open core, and hosted service. Each works for a specific shape of project.
Compression is underrated
The cheapest performance win available, ignored because it is not glamorous. Where it pays and which algorithm to pick.
The queue is the architecture
Most scaling problems are solved by making something asynchronous. Most reliability problems are caused by doing it badly.
Run the incident before the incident
Game days, failure injection, and the specific reason your untested runbook is wrong.
Python 3.15 and free-threading's second act
The GIL-free build is officially supported and the ecosystem work is the actual story. A progress report.