4 ms·

Clickbait aside, does this effectively mean we can produce latency-free games in the sense that all players could have the exact same amount of latency?

smolder 6y ago
That's already effectively possible and doesn't solve any of the problems that come with propagation delay other than keeping the effects of latency "fair".
Bjartr 6y ago
So long as the speed of light is an ultimate speed limit, the only way to ensure identical latency is to make everyone experience the same latency as the person with the worst latency since every 1000 miles of straight-line distance adds 5.4ms one-way latency under ideal conditions. So a signal following undersea cables will incur worse latency as it's not a perfectly straight line point-to-point. Better communication technologies can get us closer to that ideal, but everything we know points at the speed of light being an upper bound for any signal propagation.
lostmyoldone 6y ago
The best latencies observing "causal fairness" you can achieve is to simulate/emulate the information propagation properties of light, rescaled so that "light speed" is related to whatever time it takes to traverse the slowest link, and that all faster links are path dependently delayed as to create a similated equidistance between players. To do this anywhere close to perfectly you need absolute local time measurements with reasonable precision to determine the parameters. Anything with less latency than that will have some kind of paradoxes, as that'd essentially be the equivalent of someone transferring data (or traveling) faster than light. It's kind of odd that it turns out this way, but it seems that any general case communicating network that's paradox free must be reducible to an equivalent physical light circuit. It is still possible to make something very very counterintuitive that has a better perceived latency, is still fair for most purposes, and is at least eventually consistent. That is however more based on the relative predictability of humans over very short terms and the same of any physical simulation, and not at all causally consistent in any strict sense.

smolder 6y ago

That's already effectively possible and doesn't solve any of the problems that come with propagation delay other than keeping the effects of latency "fair".

Bjartr 6y ago

So long as the speed of light is an ultimate speed limit, the only way to ensure identical latency is to make everyone experience the same latency as the person with the worst latency since every 1000 miles of straight-line distance adds 5.4ms one-way latency under ideal conditions. So a signal following undersea cables will incur worse latency as it's not a perfectly straight line point-to-point. Better communication technologies can get us closer to that ideal, but everything we know points at the speed of light being an upper bound for any signal propagation.

lostmyoldone 6y ago

The best latencies observing "causal fairness" you can achieve is to simulate/emulate the information propagation properties of light, rescaled so that "light speed" is related to whatever time it takes to traverse the slowest link, and that all faster links are path dependently delayed as to create a similated equidistance between players. To do this anywhere close to perfectly you need absolute local time measurements with reasonable precision to determine the parameters. Anything with less latency than that will have some kind of paradoxes, as that'd essentially be the equivalent of someone transferring data (or traveling) faster than light. It's kind of odd that it turns out this way, but it seems that any general case communicating network that's paradox free must be reducible to an equivalent physical light circuit. It is still possible to make something very very counterintuitive that has a better perceived latency, is still fair for most purposes, and is at least eventually consistent. That is however more based on the relative predictability of humans over very short terms and the same of any physical simulation, and not at all causally consistent in any strict sense.