Hacker News new | ask | show | jobs
by gosub100 552 days ago
terrestrial phone / internet carried by undersea cables are gated by the relays more so than c. the ping time from US to Australia (one way) is about 115 ms (rounding down, using most optimistic data.

Light can travel over 34,000km in that time. The great arc distance from LA to Sydney is just over 12,000km. In all likelihood the fiber line connecting them doesn't follow that arc, but it shouldn't be too far out of limits. So about 2/3 of the latency is caused by relays and switching equipment.

it gets even worse for satellite, because (until starlink) communications satellites are in geosynchronous orbit, 35,000km above the equator. so talking on one means a 70km round trip, which causes its path to take over 5x more distance than the linear distance (across the surface) between those 2 cities.

3 comments

>> it gets even worse for satellite, because (until starlink) communications satellites are in geosynchronous orbit

No. There were and are other communication satellites in lower orbits. SpaceX did not invent the concept of low-orbit communications satellites. The first satellites of the 90+ Iridium constellation were launched in the late 90s. That system is still online in low orbit today. Before that there were various military/state-owned satellites. The Soviet Union and Russia were big on providing coms to areas where the geo-stationary relays could not, specifically the north, as far back as the 60s. See the Molnya program.

https://en.wikipedia.org/wiki/Molniya_(satellite)

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

Iridium existed but how much backhaul did it actually carry? it was meant for very low-bandwidth applications.
> So about 2/3 of the latency is caused by relays and switching equipment.

Not really; the speed of light in fiber optic cable is only about two-thirds of that in a vacuum. That means it takes light about ~60ms to travel the 12,000km great arc distance.

thats a great point, I failed to consider c in the propagation medium. however, OP didn't specify that either, he just insinuated that we can perceive propagation delays by c. we are both wrong (in his specific case about telecom delays, we have absolutely contrived experiments to detect it)
>> The great arc distance from LA to Sydney is just over 12km.

12,000km ?

fixed, thank you