A research house is its record. This page is the structured whole of ours: every published piece, dated, with its current standing stated plainly — so a reader arriving cold can understand what we claim, what we have computed, and what we have already been wrong about. Nothing is ever deleted. Work that is overturned or superseded moves to the suspended record below, readable in full, with the reason it fell. That is not housekeeping; it is the method.
Published June 2026 · stage0.html
The fuel half of the thesis, from the ground up: where deuterium actually sits on Mars (measured reservoir ratios, ice inventories), how the separation step works, and the honest engineering chain from landed hardware to enriched fuel — including the part a person can build on a bench. The reactor stays blocked; the fuel does not.
Published June 2026 · surface-power.html
The question everyone asks first, answered with a heat balance: sunlit orbit is power-rich but must run its radiators hot against the Earth’s glow; the Martian surface is power-poor, with a cold night sky and a warm ground and Sun that can be screened from the panel. Per square metre of panel, orbit still sheds more heat, so the surface does not make radiators lighter per watt; what it lacks is power.
Revised 27 September 2026: the claim that the thin Martian air provides a real cooling channel does not hold for radiator-sized panels; it was ten to twenty times too high. The surface’s advantage is a clean, shieldable night sky. Now backed by technical memorandum FUNNY-TM-2026-003; the original white paper stays available unchanged.
Rewritten October 2026: in the format of our other articles, with numbered and graded sources; no conclusion changed. This summary was reworded at the same time; it had called orbit cooling-poor and the surface’s advantage a cheaper mass tax, which the memorandum does not support.
Published June 2026 as the question · computed July 2026 · deltas.html
A published stellarator design taken as given and moved off Earth, parameter by parameter, across three destinations. We published the table open, sealed our expectations, computed — and report the seals we got wrong along with the ones we got right. All six deltas now carry verdicts; the machine that travels is not the machine that stays home.
Revised 27 September 2026: the same correction moves the Mars heat verdict from mild to moderate (radiator field roughly doubles) and the Mars plant from about 70,000 to about 80,000 tonnes; the Moon and Mars now match on heat.
Published July 2026 · ladder.html
What fell out of the computed deltas: the three destinations are a sequence, not a menu. Each rung is the cheapest place the next rung’s assumptions can fail — prove the machine on Earth, prove the plant on the Moon, prove the thesis on Mars. Written in conditions, never in years.
Published September 2026 · coming-back.html · with technical memorandum FUNNY-TM-2026-001
A ship that lands on Mars must make its fuel there to come home: about a megawatt, continuously, per ship. Fission carries the first several hundred ships; fusion becomes the lighter supply only above a few hundred ships per window, and only if two open conditions fall its way. Our original claim — that fusion is the return technology — was narrowed by this computation, and the page says so.
Revised 27 September 2026: the page said the external review led to seven accepted corrections; the review record shows nine. Memorandum now at Revision D (NASA context, process diagram, itemised fusion mass, calculation script published). No result changed.
Revised again 27 September 2026: the heavier Mars plant moves the fusion crossover from about 490 to about 560 ships per window; no conclusion changed.
Empty, so far — and we intend to earn additions honestly rather than avoid them. The first candidate already exists inside the work: we tested the idea that the Moon’s permanently-shadowed craters could serve as a plant-scale heat sink, and our own computation killed it — a gigawatt warms any crater out of its coldness; the cold that took two billion years to accumulate is not a resource, it is a record. The idea never reached a page of its own, so there is nothing to suspend — but that is the kind of entry this section is for, and the standard it will be held to.
When an entry does land here it keeps its full text, gains a dated notice stating exactly what observation or computation overturned it, and remains linked from whatever superseded it. Outdated is not unreadable; it is instruction.
This archive is the public face of an internal discipline: claims carry kill-conditions, expectations are sealed before computations run, and the scorecard — right and wrong — is kept. Figures throughout are order-of-magnitude systems analyses, internal physics passes June–July 2026, awaiting external review.