the successor publishes what it drops

Documentation for a late prototype in this cluster opens with a map of what it replaces. It lists what came forward from earlier rounds, what is new, and what has been set aside for a later one, and it names its predecessors explicitly enough that a reader can compare them.

The tension is that this is not how progress is usually presented. Internal records get written for people already inside, and the natural instinct is to describe the current design as though it were the design. A replacement map makes the present work look like a stage rather than a result, and a deferral list is a written record of not having done something.

The decision is to publish the debt alongside the work. Most components carried forward, a few were new, a few were set aside with their destination stated. The deferred names are the interesting part, because they can be looked up later and asked about, which is what makes the list a commitment rather than a note.

What it costs is dignity, mostly, and the permanent possibility that someone holds a deferral against you. It also leaves the record open to the reading that unfinished work is the story, when the story is that the work was sequenced on purpose.

What it proves is that the team's account of its own progress includes what it has not built, in the same document as what it has, with the gaps named rather than vague. A team that writes down what it postponed is a team whose next claim about readiness can be checked.