Hacker News Reader

Comment by guepe | Hacker News Reader

Higher throughput with higher latency. I am not strongly opinionated about tbc, but I've seen many comparisons of Las Vegas loop results compared to bus/train. Sure throughput is lower. But latency is too! And users experience latency first.

It's an interesting trade-off. They hit the target throughput. Its done. What's not to love on this project?

View on HN

Why would there be higher latency?

Tokyo, NYC and London all run subway trains during peak hours within 2 mins. That’s almost certainly better than what the Boring loop achieves.

And these are subway systems. What you would run would be an autonomous light rail system with 1-2 cabins like you see running in airports all over the world.

The latency is likely lower since you don't need to look for parking.

Slightly off-topic: The current Tokyo subway system has the unfortunate property that it turns off between midnight (~0) and early morning (~5). The Loop thing wouldn't have this likely, but I don't understand why Tokyo does this, especially for the trains that don't have a driver.

Maintenance is far easier when you don't have live trains running. If anything, NYC is an outlier for having 24 hour services, and it shows in their abysmal on-time record.

It's very, very difficult to run trains with headways of under 2 minutes (as in, under two minutes from one train closing doors to the next train closing doors). The record holder is apparently Paris Line 1 at 85 seconds during peak hours, and some lines of the Moscow Metro do 90 seconds.

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

What are the "latency" numbers you're talking about on the LV loop? Like, what is the actual metric you're measuring for that, and what's the value you've been quoted for that measure.

If you count time waiting in traffic, I bet trains have better latency...

I mean it doesn't scale in the slightest so it doesn't solve any problems really, so there's that. In terms of latency it depends, every decent underground system will have passengers waiting no more than a few minutes. Worst I saw was in Naples where I waited over 15 minutes. 15 was too long, but I'd rather wait a few (<10) minutes to ride a train than wait some amount of time less to ride a Tesla. In fact, I would wait just to avoid riding a Tesla at this point.

What do you mean by higher latency? What do you think the latency for the Las Vegas loop would be if it needs to carry 1000 passengers over e.g. 20min? Sure if there's only 5 people waiting it's quicker, but if there is only 5 people wanting to go a certain path, why do we need a tunnel at all? The road would be faster. The whole point to build tunnels is to take away congestion.

Nonsense.

If the throughput is lower, the latency is necessarily higher because you'll be waiting on the dock longer for a car to pick you up.

> What's not to love on this project?

literally all of it

A ski lift would work better.

Reply on news.ycombinator.com