| website | https://draganstepanovic.com |
It's not that the cost of code went to zero but that the cost of _plausible-looking code_ went to zero.
It might've went to zero (hint: price hikes says not) but the surrounding costs in the process (cruft, complexity carrying costs, rework, etc.) went drastically up.
It's not that the cost of code went to zero but that the cost of _plausible-looking code_ went to zero.
It might've went to zero (hint: price hikes says not) but the surrounding costs in the process (cruft, complexity carrying costs, rework, etc.) went drastically up.
Last Week in Collapse
Last Week in Collapse
"A 2,000-line PR becomes legible in a single sitting." says Linear for its new guided reviews feature.
It shouldn't be.
Any tooling that makes big changes easier or faster to review and/or deploy is still solving the wrong problem.
Reality (production) doesn't care how long it took you to review a big batch or how difficult that was.
If the batches are big and there are more of them, deploying them carries inherent risk that grows exponentially with batch size. And often you get to discover this too late in the process.
To quote Jay Forrester again: "Right intervention point, wrong direction".
