Show HN: QuadTree model for generating random road networks (opens in new tab)

Sorta. There are places that are almost completely on a grid, like Salt Lake City[0]. Even then, there is a dense downtown core, and sparser suburbs with more organic neighborhood streets further out (if not blocked by mountains), as opposed to random clusters dense roads that aren't anywhere near each other. Then there's Pittsburgh[1], whose topography prevents a large grid, and there are several colliding grids at different orientations, and the suburbs gave up.

[0] https://www.google.com/maps/place/Salt+Lake+City,+UT/@40.737...

[1] https://www.google.com/maps/place/Pittsburgh,+PA/@40.4424429...

I spent two minutes with Google Maps and managed to confirm, somewhat to my surprise, that these generated maps are actually a reasonably good approximation to areas in Nebraska.

Must make map making easier as someone actually appears to have created a physical grid system... :-)

Not an expert, but generally the further west you go in the US (until you hit the west coast, which was already settled before the big westward expansion), the more grid-like things are, especially in areas without many geographic features

as I understand it, this is because the land was more and more allocated in logical blocks -

https://en.wikipedia.org/wiki/Public_Land_Survey_System#Size...

I mean, we even have completely square or nearly completely square states -

https://en.wikipedia.org/wiki/Colorado

etc

Not explicitly, they are just dense parts of the network. But towns and cities tend to have dense network of roads. The network is represented as square grids, so it resembles a city where roads are organized into blocks - like many US cities.