Sorry SE, but endgame won't work if you don't fix something.

[Knosis]

But I think there is some truth to the fact that the problems are between users and SE, and not directly on the server.

I don't know... my latency is always 100ms or less (pinging the actually IP of the server I'm connected to) and yet I still see players moving in a "stop go, stop go" fashion as if their position is not being refreshed very often. Players still run in place before actually moving, or continue running when they have already stopped. My favorite is seeing a player take out their pickaxe before reaching their mining point, but slide on over to it. Watching players fail to run in small circles is also a good laugh as the server doesn't refresh/send their position fast enough to allow them to create a circle. So sad

EDIT: I'd love to see a video of players running around without the stop go motion that the positional delay issues in the netcode produce or players barely dodging AoE.

[Shaone]

I don't know... my latency is always 100ms or less (pinging the actually IP of the server I'm connected to) and yet I still see players moving in a "stop go, stop go" fashion as if their position is not being refreshed very often. (snip)

I think you are may be experiencing the problem I have when I don't use proxy, which is packet loss rather than actual ping delay (I also have a general ping of <100ms). So you might get the packet that tells you another player took out their pickaxe, but the packet before containing their movement to the node got lost, and then retransmitted afterwards. My ISP got me to run Wireshark and this showed a huge number of retransmissions, so it is possible to determine if this is the problem you are having.

[whilke]

I think you are may be experiencing the problem I have when I don't use proxy, which is packet loss rather than actual ping delay (I also have a general ping of <100ms). So you might get the packet that tells you another player took out their pickaxe, but the packet before containing their movement to the node got lost, and then retransmitted afterwards. My ISP got me to run Wireshark and this showed a huge number of retransmissions, so it is possible to determine if this is the problem you are having.

Their protocol runs over TCP. There is no re-transmitted afterwards. You won't get that packet for player 2 taking out their pickax until you get the packet containing the movement that was sent before it. That is how TCP works, guaranteed ordered delivery of packets. And, that is one of the major issues with people getting rubberbanding. Any small amount of packet loss on their line (pretty normal for the internet) will cause huge spikes.

The 300ms built in delay is just on top of that.