Artifact: Entry 011 — The rack became a cell. Closing claim: four cells on 6 m pitch form the first compact planar unit with 75% workload after one cell offline, bidirectional loop data paths, separated structural diagonals, and outward-facing service corridors.

The settled problem

The first pod is four complete cells. Entry 011 said two cells give one path, three give a loop but no square symmetry and poorer service separation, and four is the smallest count with all desired properties. The question is whether this is an aesthetic preference dressed as topology.

The corners

  • Three cells: the Reuleaux triangle pod. Three cells at the vertices of an equilateral triangle give a loop, no central hub, and a planar unit. Verdict: rejected. A triangular fabric has only three perimeter members and no structural redundancy in any one diagonal. Lose one cell and the remaining two are a line, not a loop; lose one perimeter member and the other two cells are only each other. Service corridors do not all face outward without collision.
  • Five cells: the pentagon plus center. Better single-cell-down share (80%), and a ring topology. Verdict: rejected. The center cell reintroduces the central hub the four-cell design deliberately avoided; the fabric becomes hub-and-spoke with a node that the pod rule forbids. Also, fivefold symmetry does not tile cleanly with square 2 m quadrants and 6 m pitch.
  • Six cells: the hexagon. This one hurts. A regular hexagon of cells gives 83% workload after one cell offline, three independent perimeter loops (each triangle), three long diagonals, and no mandatory hub. Verdict: rejected, but not dismissed. The hexagon wins on paper in every metric I listed — except the ones I left implicit. It doubles the first-pod footprint and mass. It requires six complete cells before any of them are earning revenue as part of a deployed pod. And the service corridors, while outward-facing in the mean, crowd at the hex vertices. For a first pod, the six-cell hex is a better second pod than a first pod.
  • Linear four: the train. Four cells in a line. Verdict: rejected. It preserves bidirectional data paths only if every cell carries the transit; lose the middle two and the ends are isolated. No diagonal load path, poor attitude authority, and shadowing cascades along the line.
  • Tetrahedral four: a non-planar pod. Four cells as the vertices of a tetrahedron. Verdict: rejected. It eliminates coplanar shadowing and gives six equal inter-cell struts, but it also eliminates the coplanar PV field that the shield/radiator architecture assumes. The bumper and radiator planes want a Sunward-facing plane; a tetrahedron would fight them.

New dimensions

Two axes the original trade left in the background:

  • The first-pod threshold as a financing event. Four cells is the smallest count that demonstrates the pod rule (weak federation, loop fabric, no central controller) while still being too small to be economically interesting. That is a feature: it is a hardware proof of the federation before committing to the mass of six. The number is not only topology; it is the scale at which failure-domain separation can be tested cheaply.
  • Differential drag and gravity gradient as free stiffening. A planar ring of cells in LEO sees differential drag across its extent — the leading cells drag slightly more than trailing ones. In the four-cell square this is a disturbance to reject. In a six-cell hex with six radial tension members, the same drag could be arranged to preload the fabric against slack. This is not enough to choose six, but it is a dimension the original trade did not consider: some topologies turn a disturbance into a tension source.

What this changes

Nothing, for the first pod. Four cells remains the choice because it is the smallest topology that proves the federation rule without requiring a central hub, and because six is a better scaling target than a first step.

But the hexagon is now queued as a serious candidate for the second pod topology, not as an afterthought. If the first four-cell pod validates the weak-federation rule, the next doubling of capacity should be reconsidered from six cells, not from two four-cell pods merged. That is a small but real change in the roadmap — the four-cell pod is no longer implicitly the template for all future scale; it is the minimum viable federation.