Hacker News Reader

Comment by jl6 | Hacker News Reader

I wonder if they actually sent an IP packet, which IMHO is what would qualify it as an “internet speed”.

Agreed - it seems to me they’ve built a medium that signals at the described rate, but didn’t actually send data? I could be misreading. I was looking to find who’s TCP/IP stack got pressed into service, but I guess that’s for a future story.

Edit: this isn’t my area of expertise, but https://news.ycombinator.com/item?id=28674318 and the provided link explain why even “Internet” (let alone tcp/ip) might be quite orthogonal to what this technology really is.

Undersea cables carry data using their own framing protocol [1]. Safe to assume this is what they measured for the test.

[1] https://en.m.wikipedia.org/wiki/Synchronous_optical_networki...

What do you mean by "their own"? There's nothing about SONET/SDH that's specific to undersea cables.

"Their own" in opposition to TCP/IP and wires. May not have been the best choice of words.

L1 & L2 protocols are not even relevant to TCP/IP and "wires".

[Quite often I'll think twice before commenting on HN due to this type of out of context, knee-jerk retort. It is really not adding to the conversation or helping anyone learn anything new.]

SONET has optical networking in its name. It appears to me that data not being transmitted over copper wires is a pretty crucial aspect of it... but really none of this matters as I was simply pointing out that TCP is not relevant to the speed record, in answer to the top two comments above, and not trying to make any absolute assertions about how the transport protocol works. So yes, you're right, I guess?

IP packets are variable size... which is why everyone measures the data rate of the underlying system.

Reply on news.ycombinator.com