Hacker News Reader

Comment by kayoone | Hacker News Reader

we really need a focus on low latency instead of bandwidth but i guess that's even worse in terms of marketing than upload bandwidth. But its frustrating to know that <10ms latencys are easily achievable with current technology but ISPs just don't care. Lower latency even improves web browsing a ton, voice/video calls and any kind of realtime interaction basically. Then again with 4K gaming becoming more popular, even todays bandwidth will not be enough.

View on HN

I'm not the biggest fan of ISPs but that comment isn't entirely fair on them.

In Australia the speed of light actually matters for the distances light needs to travel - this is probably the same for mid-west towns not near an AWS centre too (not sure of their presence there, but I expect they are predominately on the coasts).

The speed of light is 300,000 kms per second.

Distance wise: Perth to Sydney approx 4,000kms or 13ms travel time, and Brisbane to Sydney approx 1,000kms or 3ms travel time.

I'd say it is highly doubtful to get sub 10ms from Brisbane to Sydney if you're going via even just a handful of routers. And Perth to Sydney? Forget it!

I'm ~2 light-ms from us-east-1, and my latency is 49ms. My latency is 11 times what it would be if there were a direct fiber run from my laptop to us-east-1.

Just for nitpicking, light doesn't travel the same speed in fiber as in vacuum. AFAIK the refraction index of the core is typically ~1.45 so your latency is "only" around 8 times larger than the theoretical limit.

If you want to be fair to ISPs, don't look at cases where they fail to do the impossible or very hard, look at cases where they fail to do the very easy.

The standard deviation of any ping test I do is at least 3ms due to how DOCSIS works, and I can't ping my next door neighbor (same ISP) through the internet in under 16ms. I've done traceroutes against several servers that are in the same city as I am in North Carolina and the only one where my traffic didn't first go to Atlanta or DC (or both) was to a server hosted by an ISP that has no physical presence outside of North Carolina. My cable modem and my ISP's CMTS each have out of control bufferbloat that adds hundreds of milliseconds of latency under load in each direction, which can't be entirely mitigated by my router's traffic shaping and AQM. There's little reason to believe that they've got got any AQM further upstream given the large latency spikes I see even when my last-mile link is quiet. Disregard for latency is pervasive in the design of the ISP's network.

Sure, Australia is super huge and <10ms is a different thing. I live in germany and there are ISPs where you can get <10ms in the whole country. Sadly when i moved to Berlin the ISP situation got worse than in my previous much smaller city and i am "stuck" with 30-40ms latency. I somehow care a lot about latency, but i can live with it ;) I have seen some people with Fiber connections basically get <5ms everywhere in the country, so the tech is there, we just have to get the mainstream interested. At the current rate ISPs will squeeze the last drops of bandwidth out of copper and cable connections until finally moving to fiber, but it's understandable as the mainstream demand isn't there and getting Fiber into all homes is a massive undertaking.

You're right, but it's even worse than that. The speed of light in optical fiber is closer to 200,000 km/s.

Reply on news.ycombinator.com