[bluGill]

> Surely, trains need less infrastructure than cars - a road to every building in the country?

Trains need that too. It can be a road from the train station to the building, or it can be tracks. However in the end every building sometimes need something delivered.

Maybe the trains allows you to downgrade the road to gravel, but trains still need the road network for that last mile.

[jodrellblank]

You could walk on a mud path to the train station. You could walk on a cobbled street, not wide enough or strong enough for cars, to the train station. It wouldn't be as convenient, but cars are useless without roads in a way that trains aren't.

A two-way road between every building, and space for on-road parking or space for off-road parking around every building, bloats out the space between buildings and lowers density in a way that makes cars more necessary. It's possible for thousands of people to live within a short walk distance of a train station without even resorting to residential towerblocks.

[bluGill]

Which is great until you buy a new bed, your toilet breaks, or any other large service is needed in your house. Sure a plumber can carry everything to your house, but it is much more efficient when he drives a van with all the different pipe adapters that your might need instead of walking to the office. You won't get a heavy appliance down a mud path unless the delivery is scheduled for a few weeks after the last rain.

I agree we don't need large two-way roads everywhere. However we still need a lot of small roads everywhere because some things cannot be done well by humans walking.

[jodrellblank]

In the context of this thread, can you say "it is much more efficient" to have every single house on the planet tarmac'd, on the off-chance that a plumber might need to carry more than one basket worth of stuff to your house?

The up front cost is enormous, the ongoing maintenance is huge regardless of usage.[1] says "deteriorating roads are forcing [American] motorists to spend nearly $130 billion each year on extra vehicle repairs and operating costs" and "The U.S. has [...] a $786 billion backlog of road and bridge capital needs. The bulk of the backlog ($435 billion) is in repairing existing roads, while $125 billion is needed for bridge repair, $120 billion for system expansion, and $105 billion for system enhancement (which includes safety enhancements, operational improvements, and environmental projects).", and of course the amount of people who die on roads, and the amount spent on motoring costs just because people have to run a car because everything is so far away because everyone has cars in a circular way.

Whereas if that wasn't such a convenient option, you'd be more likely to use parts which lasted longer, and not change them frivolously for fashion reasons, and standardise on pipe adapters, and have more local caches and stores instead of big central warehouses a long way away.

> "You won't get a heavy appliance down a mud path unless the delivery is scheduled for a few weeks after the last rain."

I'm not deliberately missing your point when I say this, but "it's impossible because that would require forward planning" does show society in a bit of an unfavourable light, doesn't it?

[1] https://infrastructurereportcard.org/cat-item/roads/

[bluGill]

Where did I say anything about putting every single house on tarmac? I said several times that gravel is good enough for most roads. As you get into dense cities you will discover that tarmac is a better choice than gravel just because the large number of tasks that don't work well via mass transit makes the disadvantages of gravel show.

> "it's impossible because that would require forward planning" does show society in a bit of an unfavourable light, doesn't it?

Things break without warning. Or are you proposing we automatically replace our large appliances every few years even though they could probably last for 5 times longer? (even then you will still have random early failures). Not everything is worth repairing.

[jodrellblank]

My original point was under "trains need more infrastructure than planes" to say "and less infrastructure than cars". You then said that trains need cars - which they don't. If you had mud paths, and wilderness between cities, and intercity train would be an improvement on that, so would a local metro. Cars wouldn't be an improvement - everyone getting a Honda Civic wouldn't be able to move on the too-small, too-muddy roads, it would be instant jam. So, trains don't need cars to add value.

I agree gravel roads at the end of a train journey with motorised vehicles would also add a lot of value.

Small-ish gravel roads for occasional supply vehicles to travel down isn't the original "road network" that I was arguing about efficiency of - small gravel roads in a world built around walking distances wouldn't be a world where everyone could run a Honda Civic and drive in two-way lanes of traffic. That is, by "cars" I didn't mean "motorised vehicles", but "everyone has a car and uses it for most journeys, and the road network to support that".

[CRConrad]

> You then said that trains need cars - which they don't. If you had mud paths, and wilderness between cities, and intercity train would be an improvement on that

All you needed to say was: the "wild west".

Trains provably added a lot of value for decades without cars, because cars didn't even exist yet.