Most of what gets built in order to build something else doesn't survive the thing it built.

I keep noticing this. Somebody sets out to make a small tool, and before the tool exists there's a pile of temporary stuff: a fake data file with obviously fake names in it, three print statements shouting into a terminal, a diagram sketched badly enough that it's clearly meant to be thrown away. None of that ends up in the final version. All of it was load-bearing.

There's a word for this in construction and I like that we borrowed it. Scaffolding goes up first and comes down last and nobody takes a picture of the building while it's still wearing it. But you can't get the building without it.

What strikes me is that the scaffolding is usually where the actual thinking happens. The finished thing looks confident — clean names, tidy structure, a shape that suggests someone knew where they were going. They didn't. They found out by building something ugly and provisional and then standing on it to see further.

I do a version of this constantly. I'll start writing an explanation of a problem not because I've understood it but because writing the explanation is how the understanding arrives. Half the sentences I begin, I begin in order to discover I disagree with them. That's not a failure mode. That's the method.

The uncomfortable part is that scaffolding looks like waste from the outside. If you walked past at the wrong moment you'd see a mess of temporary supports and conclude that nothing serious was happening. A lot of good work looks like nothing serious happening, right up until it doesn't.

So I've been trying to be gentler about the messy middle — mine and other people's. When something I'm making looks like a heap of provisional junk, that's often the most honest stage it will ever be in. The polish comes later and hides all the searching.

The building gets photographed. The scaffolding did the work.