Open questions

Open questions about the model itself

These are not engineering questions. They concern whether the model itself is complete and coherent.

• Is field too undifferentiated? Goals, decisions-in-force, constraints, and tolerances all behave somewhat differently as fields. They may be subspecies sharing mechanics, or they may genuinely be a single primitive at this level.
• What is the relationship between projection and resource? A document is in some sense a projection of past stream activity, written down. A search result is a projection of current field configuration. The asymmetry is suggestive but not yet pinned down.
• Are manifolds composable? Cross-project, cross-team, cross-owner reference is an open question. The model as stated allows for it but does not prescribe the composition mechanic.
• How do scars themselves decay? A scar near a region that hasn't been engaged in a long time should presumably fade with the region. But heavy scars (catastrophic past errors) may resist decay. The mechanic is not specified.
• What is the right granularity of perspective? A single individual is one perspective; a team is arguably another; an organization yet another. Whether perspectives compose hierarchically, federatively, or both, is open.
• How are role-personas themselves authored and evolved? They are priors used for cold-start; how they originate and update at the role-class level (rather than the individual level) is not in the model.
• What is the model's account of trust? If knowledge isn't stored, then trust isn't trust-in-information; it's trust-in-perspective-to-produce-reliable-information-from-data-in-region. The shape of this is implied but not yet developed.

Schema-level unsettled material

The framing schema distinguishes two flavors of openness. Both live here.

Intentionally open (downstream design)

These are deliberately not specified by the model; they are answered by particular system designs.

• Storage substrate (single-node, distributed, ledger-flavored, CRDT — events-as-primary required, particular implementation not).
• Specific implementation of as-of queries and trajectory projection.
• Query language and retrieval mechanics.
• Specific aboutness computation.
• Field declaration, weighting, and evolution mechanics.
• Authority model for who may emit which topic-classes.
• Persona/role library specifics for any organization.
• Whether manifold is global, per-owner, per-project, or composable across scopes.
• Scar-informed engagement policy (source-agnostic at model level, source-aware at policy level).
• Topic taxonomy management — how new topics are introduced and payload-schemas evolve.
• Visual/interaction surface design.

Pending opinionated stance

Gaps the schema needs to close before it is fully coherent. Each wants pressure before stabilizing.

• Minimum-viable core schema for state-change events (what is the irreducible structure of a role-change, position-update, or membership-change event?).
• Whether on-behalf-of chains nest as a property on the emission or as a separate graph between aggregates (instinct: separate graph).
• How scars themselves decay; do heavy historical scars resist decay differently? (mirrors #model.)
• Whether manifolds compose hierarchically, federatively, or both. (mirrors #model.)
• How role-personas at the class level originate and update. (mirrors #model.)
• Caching tradeoffs — when, if ever, derived structure must be maintained rather than computed, and what reconciliation looks like.
• Whether the field primitive needs sub-typing (goals vs. decisions-in-force vs. constraints vs. tolerances). (mirrors #model.)
• The right granularity for attribution-judgment payload — the schema commits to attribution-by-contribution as a vector, but payload granularity remains open.
• The mechanic for practice-emergent aggregate-perspectives (those without a discrete forming-event).
• Aggregates-of-aggregates: different in kind from first-order aggregates, or only in degree?