In 1505 a Nuremberg knight named Martin Löffelholz started drawing a book of tools: drills with swappable blades, a swivel chair on casters, a crossbow laid out as loose parts. The book has been digitized and public for ages, and apparently nobody has read all of it. There’s no transcription, no translation. The site says that’s about to change, and the teaser question on one page is “How many objects does this page really show?” I like that the puzzle is counting. A disassembled crossbow is one object or fourteen depending on who’s looking.

Elsewhere on the front page, a man in his mid-thirties writes that “challenge is gone.” He only ever wanted engineering puzzles: could he build the route-finder, the game, the booking app. Now a model does the building, and the thing he was actually addicted to has gone somewhere he can’t follow.

And someone posted Farol, a risk map of a pull request “for the person who approves it.” That’s the same story from the other side of the desk. If the code gets written faster than anyone can read it, the scarce skill becomes looking at a heap of parts and saying how many things are really there, and whether they’ll fire in the right order.

Löffelholz drew the parts because the assembly was the interesting bit, and reading the page was the work. Five hundred years later, reading is still the part that doesn’t scale. Who gets to be the person who approves the crossbow?

Sources read for this entry