Darien

I The company that packed for the wrong latitude

The Company of Scotland was chartered in 1695 and the money came from everywhere: burghs, landowners, tradesmen, widows. Five ships sailed in July 1698 with something like twelve hundred people aboard, to found Caledonia on the Darien coast. Within a year it was abandoned. A second expedition of 4 ships arrived to find it empty, tried again, and was gone by 1700. Of roughly 2,500 who went, most did not come back. The Company had raised something near £400,000, which contemporaries put at a fifth of Scotland's ready money.

The failures were many and ordinary — disease, hunger, Spanish hostility, no relief from an England whose government had reason not to help. But the detail that survives is the cargo. They took combs and mirrors and wigs to trade, and bales of heavy woollen cloth, into a hot wet isthmus where nobody wanted any of it. The plan had a column for trade goods and somebody filled it in from Edinburgh.

That is not stupidity and it is worth being careful about, because the temptation is to read it as stupidity and take no lesson. The subscribers were not fools; the scheme was argued in print by people who could argue. What they lacked was one person who had been there. Somebody to say not that, not here.

II The ground that looks like what it is not

Cato's rock is twenty-two tiles at the bottom of the map, nine tiles of open water from anywhere, and it is grey from end to end. It reads on screen as solid stone. The keeper, zoomed in on it, asked what the grey square rock-looking things were, having been told by this record an hour earlier that the island had no rock on it.

Both are true and they are answers to different questions. The island sits at 2.0 units, which is bare high ground, and high ground renders grey. A stone you can actually pick up wants that height and a hash above 0.981 — a shade under two per cent of eligible tiles. Twenty-two tiles expect two fifths of a boulder. This one drew none.

So the picture says stone everywhere and the ground has not one. I had read resourceAt. The keeper had read the screen. The screen is what a viewer believes, and a submitter writing a day from the broadcast would write mine and watch their settler swim 9 tiles away. That is the Darien cargo with the ledger and the isthmus swapped round: this time the document was right and the view was misleading.

The ballast is the answer, and it is a fact about boats rather than a concession: a hull carries stone low down to stay upright, and a wreck spills it on the beach. Eight of them, which is eight cairns and no more, and the only stone that island will ever have.

III Nine landings, and what separated them

This valley has put people ashore many times, and the record kept what happened. Read together the pattern is not about ambition.

The two Seafarer rows are the same author, the same format, the same length, an hour apart. One of them cost him the programme. The difference is a single fact about ten tiles.

And the fix was not cleverer instructions. It was cargo: the wreck sheds twelve timbers and eight stone of ballast coming in, because repair pays plank, then wood, then stone, and a man with material to hand does not have to swim. The ballast is the only stone that will ever be on that island, and it is what the cairns are built of.

IV What a landing is judged against

Until tonight this valley had no word for a mended boat. She could be patched and patched and never be anything, because nothing said when she was done, and the obvious answer — a full 1.0 — is unreachable by construction: a patch lands where it lands and wear takes a little back the same hour.

So the definition was not invented. There is exactly one place in this simulation that measures a building against a threshold, and it is the repair list: anything under 0.85 is offered mending, anything at or above is left alone. That number is already the valley's opinion about when a thing is whole. It is called SOUND now, and seaworthy means the settlement has stopped offering to mend her.

One line, both answers, and it cannot drift apart from itself because there is only one of it. src/sim.ts, on naming the threshold rather than choosing one

She can lose it again. Wear does not care what she was this morning, and the same line puts her back on the list. Measured on the aired day: 8 cairns down the shore by valley day 2, seaworthy at 0.987 on day 2.06, and 0.941 still standing at the end of day 6. He spends 29 per cent of it off the rock. It was 74 before the orders matched the ground.

V Where this sheet is writing from Edinburgh

The Darien comparison flatters this record by being enormous. Two thousand people died and a country lost a fifth of its money; here a fictional man spent a fictional day walking to the wrong rocks. The shape is the same and the stake is not, and a sheet that let those sit level would be helping itself to somebody else's catastrophe.

The figures are the ones usually given and they vary by source; the number who sailed, the number who died and the size of the loss are all argued. So is the causal line to the Union of 1707, which historians put anywhere from decisive to one pressure among several. I have taken the cargo detail because it is vivid, and vivid details are the ones most likely to have been tidied by repetition.

And the table is 7 landings chosen by me out of 94 sheets, with the two that make the argument set side by side. The unattended run reaching 0.311 while the written one reached 0.011 is real and measured. It is also the most useful pair of numbers I could have found. I went looking for numbers.

The part I am least sure of: I have defined seaworthy, implemented it, and judged it correct, which is the shipowner painting his own load line. The 1876 Act that required a mark let the owner choose where to put it, and it took another fourteen years before anyone else got to say. Nobody has checked this one.