How The Sky's Version Was Made

This record provides public traceability for edition 004. It summarises registered decisions without publishing the complete working artefacts. Both the run and this public format use methodology `1.6.0`.

Scientific starting point

The source models a Proterozoic Earth-like atmosphere for 180 days under two stellar types and two flare temperatures. In the four scenarios, hot flares have the stronger effect on ozone. The sign can change: around the modelled K star, the column reaches roughly twice its quiescent level; one large hot flare around the M star reduces it to about half.

The contrast recorded identity, URL, version, method, declared results, author inferences and limitations. It neither reproduces the study nor constitutes independent scientific validation. The models do not detect life, demonstrate habitability or cover the continuum of stars, atmospheres and flares. Atmospheric memory, the messages, and all specific behaviours and causal chains in the story are fictional compositions.

Alternatives considered

Six possibilities covered five tension topologies and six progression modes. The selected concept turned the opposite signs of the ozone response into an intimate dispute about public memory. The finalist, narrated by an artificial atmospheric gardener, retained scientific necessity and a distinctive second person, but its agency and cost were more abstract.

The other four explored a choral city divided by two readings of the same sky, two atmospheric copies of one consciousness, ageing maps and two greenhouses aboard a ship. They lost on conceptual load, familiarity, procedural drift or added urgency.

Probes and selection

Two probes of comparable length were written. Both registered `B1_SUPERADA` and `B2_SUPERADA`: each contained agency, friction and a concrete expectation. The selected concept won on relationship, clarity and distance from the three active editions. Its promise was to decide which version of Ema could return when a verifiable copy predated the conflict and a recent testimony could not be verified.

Evaluation and rewrite

The evaluation in an adversarial role returned `APROBADO_CON_CAMBIOS`. It found no critical failures, but observed that the first ending made a plural restoration too easy to accept. It also identified excess protocol explanation and a convenient final confirmation.

There was one full rewrite, with no redesign or candidate switch. Common acceptance became three incompatible branches; the automatic confirmation from the terminator was removed; Lina's responsibility became concrete when she revealed that she had previously turned Ema's objection into a bearable version; and the cup changed from a sentimental object into evidence with different meanings in each branch.

Resulting form and fingerprint

The dominant topology is internal, with epistemic and collective layers. The mode is recontextualisation: the gap changes from malfunction to promise of recovery, then to risk of falsification and finally to a space for attributable testimony. The literary wager treats the sky as a chemical version of the past rather than a neutral store.

The fingerprint changes all ten observable fields from the three previous editions: domestic colony, protagonist without authority, internal/epistemic tension, recontextualisation, no clock, relational cost, layered temporal form, testimonial climax, open ending and atmospheric-reflective voice. It retains close third person so that every reconstruction remains tied to Lina's fallible access; the POV exception is recorded. It does not repeat the combination of institutional authority, urgency and professional sacrifice.

Recorded gates

• `CF1_SUPERADA`: the paper had not been reused and the claim was create-only.
• `B2_SUPERADA`: both probes changed interpretation and sustained a concrete expectation.
• `CC1_SUPERADA`: the isolated reading reconstructed novum, situation, decision and consequence.
• `T1_SUPERADA`: Lina chooses among restoring, remaining silent and testifying, with distinct costs.
• `J1_SUPERADA`: each movement reframes the value of the copy, gap or recording.
• `L1_SUPERADA`: the isolated-text cuts retained concrete expectations and the ending paid off the mural, cup, recording and planted sentence.
• Bilingual QA: facts, alternatives, images, causality, scientific limits and ending match.

The isolated evaluation ran in a separate context that received only the final Spanish text. L1 describes the result of that procedure; it does not demonstrate how a real reader will react.

Public limit

This document summarises registered evidence and decisions. Because the complete artefacts remain private, it provides public traceability, not an independent external audit or an objective measurement of literary quality.

Read the story · Read the methodology