A short piece of fiction, written by API Evangelist. The reader is imagined; the post, and every number and quotation from it, is real: "A Common Schema For APIs, Distilled From What Providers Already Publish," Kin Lane, November 11, 2026, https://apievangelist.com/2026/11/11/a-common-schema-for-apis-distilled-from-what-providers-already-publish/
Marisol is an integration engineer at a small freight software company, and she reads the post on a Wednesday afternoon while a build runs. She opens it expecting a standards pitch, and she is already tired of those.
The first paragraph does not read like one. Kin Lane says he doesn't want another standard handed down by a committee. The rule he sets himself is "distil, never invent", and she likes that on instinct. Her team has spent two years arguing over field names in meetings, and the idea of measuring what providers already publish feels more honest than another meeting.
Then she hits the numbers: 207 domains measured, 129 built, 6,543 provider crosswalks. She notices she is suspicious of them. A property counts as core when half the providers with a readable schema declare it. Half is a lot, but also an arbitrary line. She says as much to the empty room. Why half? Why not forty percent? The post does not claim the line is sacred, but she is not sure the author would defend it in a room full of people who have to live with the result.
What softens her is the section on what the method cannot do. The post says it only sees what providers publish as OpenAPI in the catalog, and that a company implementing a standard without publishing a spec is invisible. Her own company is in that group. Customers get a PDF and a sandbox key, and the spec stays internal. She feels a small, unreasonable sting at being invisible, and then a worse one, at realizing her company has never shown up in anyone's measurement of anything.
She goes back to the standards section, since ocean shipping is her world. The table lists DCSA with three providers that implement it and four that name it in a spec. She had assumed everyone in her corner followed it. The line that stays with her is "Naming a standard is not implementing it." She thinks about her own integration docs, which mention DCSA in two places. She is not sure whether that counts as implementing it or just name-dropping, and she doesn't enjoy not knowing.
She also catches herself disagreeing with the post's tone. It is calm about a gap that, for her, has a cost. Every mismatch between her field names and the common ones is a mapping some customer's developer writes by hand. The author calls it a finding, and she thinks of it as someone's Friday evening.
Still, the post ends with an invitation, and it is small. If you publish an API in one of these domains, open its crosswalk and look at how your names map to the common ones. She notes that the author is careful to say "A crosswalk is a mapping, not a grade", which makes it easier to click.
She doesn't write a manifesto or start a migration, and she doesn't forward the post to her manager with an opinion. She opens a blank note and lists three of her company's date and status fields, the ones she suspects drift the way created and createdAt do. Next to each, she leaves a space. Tomorrow she will look up the ocean shipping crosswalk and fill in the spaces, and only then decide whether she has anything to argue about.
The build finishes. She leaves the note open.