Because Wizard said it was not possible

Because the distance is great enough that it has an impact.
The undersea fibre has repeaters every x amount of km ( im not sure how many km ) infact the biggest part of the undersea cable is the power lines for all the repeaters on the line. Everytime a repeater is "hit" latency gets added. All of these add up together to add abase minimum that you can expect. This does not even account for traffic shaping and congestion.....

The undersea repeaters are all-optical repeaters - unless I'm mistaken they commonly use EDFAs. The power lines are from the landing points, and yes, that is copper. But they merely feed the repeaters, they don't contribute to the latency. I'm not sure about latency in an EDFA, it's been years since I've studied those. But all-optical doesn't add as much latency as optical-electrical-optical conversion does.

In networking, the main cause of latency is processing (unless you're working with satellite :-P ), not propagation. Switch off all layer 2/3 processing on a device, and its latency generally decreases.

Light in the fibre does not move at the speed of light (in a vacuum). The fiber have a high density slowing the light down + the light does not move in a straight line but bounces off the walls.

True, but let's say your 43ms is correct. That is much less than 200ms. What I'm trying to say is that the long distance propagation is not the main contributor to the latency.
 
True, but let's say your 43ms is correct. That is much less than 200ms. What I'm trying to say is that the long distance propagation is not the main contributor to the latency.

Not sure the actual length of the cable to Paris, but I suspect it close to double the length I used in my calculation. 100ms thus being the distance limiting latency while the repeaters, converters, etc uses another ~100ms. We had recorded latencies of about 150ms before, but that was prior to World of Warcraft and the huge uptake in line capacity. Can ask some of the old diablo2 players here that still played on EU servers.
 
Don't forget that you guys are calculating a one way trip. You need to double that to measure your latency to and back.
 
Don't see why the big whoohaa, I and most fellows raiding with me on WA + a Tunnel service raid with about ~220ms. 8ms Higher than the screenshot, but not surprising since I'm in George.
 
averaging about 230-300 on WA shaped with a basic tcp/udp registry hack to increase amount of packets sent and received which seemed to have a small effect. Looking into a tunneling service though... tried gamepath but the testing username and password doesnt work?

Maybe i should just pay the money and get a month account to test. is it worth it?
 
how much of a difference does it make? I presume it is only really applicable if your latency is around 500/600. if its around 300 it cant really improve it by much.

If you run without it, what is the difference in latency?
 
My difference in latency without it varies, can be about 330-450 without, with its 210-280. Early in the morning i get 180
 
Top
Sign up to the MyBroadband newsletter
X