Most OpenAPI editors aren't doing what I think they should. They focus on two things: linting the OpenAPI document and rendering an API reference preview. The second thing puts the focus on creating good-looking API references. API creators will, naturally, try to write their OpenAPI documents so that the API reference looks nice. When you're creating, you want to see a preview of what you create. When you write code, you want to see the output of the code running. When you write a markdown document, you want to see it rendered—either in place or on a new panel. When you write an OpenAPI document, you want to see the API it defines running. What you want is to feel the result of what you're creating.
You mean like an actual mock or something like that?
yeah, an interesting vision, had that feeling 10 years ago, using apiary (not even aligned with the emergent standards that one) and swagger ... but I haven't played with these tools for a long time now you mean, the devx is still that much out of sync? "Most OpenAPI editors aren't doing" but some are? can you give examples?
“When you write code, you want to see the output of the code running”… so u have to press “run”… in api design that is press “run mock”….
Definitely agree. But there’s a huge disconnect still between the process of defining an API and what it delivers. And how to keep the two aligned.
Very much in agreement with this.
Like a rudimentary implementation?
You can run Prism in watch mode!
Hell, even the openapi maintainers don't do what you think they should due to their poor understanding of apis https://medium.com/@apiexpert/openapi-openly-humiliates-themselves-82e4a30a4899
Principal API Architect at Microsoft and HTTP API specialist
1moBut what does an API look like?