AurOn2COOKIES! FREEDOM, AND BISCUITS!AustraliaJoin Date: 2012-01-13Member: 140224Members, Forum Moderators, NS2 Playtester, Forum staff
edited March 2013
not at all.
Servers dont hang onto packets (data) all the way to the recipient, they only dot in co-ordinants and then send the packet, which is no different from person to person.
Only difference is how fast they recieve them.
Think of it as sending mail, do you take longer in the post office if you are sending mail to the next town, or the next country? Nope, no time difference.
If NS2 were ported to xbox, then yeah. Pay an annual fee for peer-to-peer networking, and then get ads embedded in your home screen! It's the future of gaming.
Sorry, but to answer the question, I don't believe so, I can't prove this, only the developers can prove this, but when my ping is 3000, everyone else remains at 30-60.
not at all.
Servers dont hang onto packets (data) all the way to the recipient, they only dot in co-ordinants and then send the packet, which is no different from person to person.
Only difference is how fast they recieve them.
Think of it as sending mail, do you take longer in the post office if you are sending mail to the next town, or the next country? Nope, no time difference.
The server has to rollback to the perceived point in time of the player. This means the server has to have save the state (aka frame) for each tick and then interpolate between two known frames. Those euro players are going to be causing access to frames from a different state than most every other player - which means slower lookups due to cache misses. The more exceptions the server has to compensate for, the slower the process. This doesn't even touch on problems like dropped packets from unreliable protocols and the extra work it entails.
The penalty for such may be negligible to detrimental - guess which is most likely for a CPU hungry game like NS2 that's already straining from running interpreted code?
The server has to rollback to the perceived point in time of the player. This means the server has to have save the state (aka frame) for each tick and then interpolate between two known frames. Those euro players are going to be causing access to frames from a different state than most every other player - which means slower lookups due to cache misses. The more exceptions the server has to compensate for, the slower the process. This doesn't even touch on problems like dropped packets from unreliable protocols and the extra work it entails.
Too much misinformation in this post to know where to begin......
matsoMaster of PatchesJoin Date: 2002-11-05Member: 7000Members, Forum Moderators, NS2 Developer, Constellation, NS2 Playtester, Squad Five Blue, Squad Five Silver, Squad Five Gold, Reinforced - Shadow, NS2 Community Developer
not at all.
Servers dont hang onto packets (data) all the way to the recipient, they only dot in co-ordinants and then send the packet, which is no different from person to person.
Only difference is how fast they recieve them.
Think of it as sending mail, do you take longer in the post office if you are sending mail to the next town, or the next country? Nope, no time difference.
The server has to rollback to the perceived point in time of the player. This means the server has to have save the state (aka frame) for each tick and then interpolate between two known frames. Those euro players are going to be causing access to frames from a different state than most every other player - which means slower lookups due to cache misses. The more exceptions the server has to compensate for, the slower the process. This doesn't even touch on problems like dropped packets from unreliable protocols and the extra work it entails.
The penalty for such may be negligible to detrimental - guess which is most likely for a CPU hungry game like NS2 that's already straining from running interpreted code?
The difference is neglible on the server; every player is running in his own universe anyhow, so the difference between a low-ping and a high-ping player is only in the number of saved network states that are kept around in memory.
There is a slight extra CPU cost on the client for running in high latency (due to having to predict more moves), though 225 removed most of that cost by running the prediction in its own thread.
matsoMaster of PatchesJoin Date: 2002-11-05Member: 7000Members, Forum Moderators, NS2 Developer, Constellation, NS2 Playtester, Squad Five Blue, Squad Five Silver, Squad Five Gold, Reinforced - Shadow, NS2 Community Developer
If my ping goes up to 600 I can really see the client performance drop.
I'm guessing you have a dual core? The prediction thread competes with the logic thread and render thread; at 600 ping, NS2 probably needs more than 2 cores to run all of its threads flat out.
If my ping goes up to 600 I can really see the client performance drop.
I'm guessing you have a dual core? The prediction thread competes with the logic thread and render thread; at 600 ping, NS2 probably needs more than 2 cores to run all of its threads flat out.
Comments
Servers dont hang onto packets (data) all the way to the recipient, they only dot in co-ordinants and then send the packet, which is no different from person to person.
Only difference is how fast they recieve them.
Think of it as sending mail, do you take longer in the post office if you are sending mail to the next town, or the next country? Nope, no time difference.
Sorry, but to answer the question, I don't believe so, I can't prove this, only the developers can prove this, but when my ping is 3000, everyone else remains at 30-60.
Lagging players make every1 med (is only a gaem), then they leave and performance goes up.
The server has to rollback to the perceived point in time of the player. This means the server has to have save the state (aka frame) for each tick and then interpolate between two known frames. Those euro players are going to be causing access to frames from a different state than most every other player - which means slower lookups due to cache misses. The more exceptions the server has to compensate for, the slower the process. This doesn't even touch on problems like dropped packets from unreliable protocols and the extra work it entails.
The penalty for such may be negligible to detrimental - guess which is most likely for a CPU hungry game like NS2 that's already straining from running interpreted code?
Too much misinformation in this post to know where to begin......
The difference is neglible on the server; every player is running in his own universe anyhow, so the difference between a low-ping and a high-ping player is only in the number of saved network states that are kept around in memory.
There is a slight extra CPU cost on the client for running in high latency (due to having to predict more moves), though 225 removed most of that cost by running the prediction in its own thread.
I'm guessing you have a dual core? The prediction thread competes with the logic thread and render thread; at 600 ping, NS2 probably needs more than 2 cores to run all of its threads flat out.
Yeah I have a 3GHz Core 2.