3 ms·

> That's clearly stretching the laws of physics, at the very least This is not clear to me. Few mobile users have pings to anywhere pushing up against the spee

by rrss 7y ago

> That's clearly stretching the laws of physics, at the very least This is not clear to me. Few mobile users have pings to anywhere pushing up against the speed of light, and the bandwidth/loss/routing is not close to being limited by physics.

zzzcpan 7y ago
No, he's correct here, it's an extra indirection and so overall it can only hurt performance, not improve it. For example, ISPs often have caching servers from Google installed and Google is the biggest traffic generator. If you route traffic through somewhere else you are going to reach different Google's caching servers that are farther away and over more congested links.
- CharlesW 7y ago
  > No, he's correct here, it's an extra indirection and so overall it can only hurt performance, not improve it. Really? The cost of an extra hop is just one of tens (hundreds?) of factors that would determine if using WARP would be faster for a particular scenario.
  - zzzcpan 7y ago
    That's not an extra hop, but essentially two different paths with many different hops instead of a more direct one. I'm sure there will be edge cases where this can be faster due to one of those factors, but for most destinations it will be slower.
- Pfhreak 7y ago
  In aggregate, is it necessarily an extra indirection? If, for example, your ISP bounces around a bunch getting to your destination and Cloudflare has a faster path, isn't it possible they have the faster path?
xorcist 7y ago
It was not meant in a literal sense, I hope that much was clear. The language was suggestive on purpose considering the article being commented on. A longer path takes longer time as speed of light varies comparably little, that much is physics. And sending traffic by way of a third party is very likely to make for a longer path, for most people in most circumstances. Not by necessity but because few have that lousy routing agreements, unless we're talking about special circumstances such as 6to4 tunnels and the like.

by rrss 7y ago

zzzcpan 7y ago

No, he's correct here, it's an extra indirection and so overall it can only hurt performance, not improve it. For example, ISPs often have caching servers from Google installed and Google is the biggest traffic generator. If you route traffic through somewhere else you are going to reach different Google's caching servers that are farther away and over more congested links.

CharlesW 7y ago
> No, he's correct here, it's an extra indirection and so overall it can only hurt performance, not improve it. Really? The cost of an extra hop is just one of tens (hundreds?) of factors that would determine if using WARP would be faster for a particular scenario.
- zzzcpan 7y ago
  That's not an extra hop, but essentially two different paths with many different hops instead of a more direct one. I'm sure there will be edge cases where this can be faster due to one of those factors, but for most destinations it will be slower.
Pfhreak 7y ago
In aggregate, is it necessarily an extra indirection? If, for example, your ISP bounces around a bunch getting to your destination and Cloudflare has a faster path, isn't it possible they have the faster path?

xorcist 7y ago

It was not meant in a literal sense, I hope that much was clear. The language was suggestive on purpose considering the article being commented on. A longer path takes longer time as speed of light varies comparably little, that much is physics. And sending traffic by way of a third party is very likely to make for a longer path, for most people in most circumstances. Not by necessity but because few have that lousy routing agreements, unless we're talking about special circumstances such as 6to4 tunnels and the like.