Pages: 1 :: [one page] |
|
Author |
Thread Statistics | Show CCP posts - 0 post(s) |
Kitten Empress
Ametat Security Amarr Empire
391
|
Posted - 2013.06.09 05:22:00 -
[1] - Quote
https://developer.valvesoftware.com/wiki/Source_Multiplayer_Networking
See, that's an engine from 2004 that does networking better than CCP. Can't you put in that kind of lag compensation?
"All source code for lag compensation and view interpolation is available in the Source SDK. See Lag compensation for implementation details.
Let's say a player shoots at a target at client time 10.5. The firing information is packed into a user command and sent to the server. While the packet is on its way through the network, the server continues to simulate the world, and the target might have moved to a different position. The user command arrives at server time 10.6 and the server wouldn't detect the hit, even though the player has aimed exactly at the target. This error is corrected by the server-side lag compensation.
The lag compensation system keeps a history of all recent player positions for one second. If a user command is executed, the server estimates at what time the command was created as follows:
Command Execution Time = Current Server Time - Packet Latency - Client View Interpolation
Then the server moves all other players - only players - back to where they were at the command execution time. The user command is executed and the hit is detected correctly. After the user command has been processed, the players revert to their original positions. "
Really source engine games seem to be the most stable for me even in high pings. I can get 150 ping and not notice a thing. |
xxwhitedevilxx M
Maphia Clan Corporation CRONOS.
338
|
Posted - 2013.06.09 05:31:00 -
[2] - Quote
i have something like 450 ms on US battle server, so it's useless anyway :( |
Kitten Empress
Ametat Security Amarr Empire
392
|
Posted - 2013.06.09 05:56:00 -
[3] - Quote
xxwhitedevilxx M wrote:i have something like 450 ms on US battle server, so it's useless anyway :( The lag compensation is apparently effective up to a second, or 1000 ms. You won't get a perfect experience, but if you shoot someone you will hit them at least. |
Sontie
Ill Omens EoN.
637
|
Posted - 2013.06.09 06:14:00 -
[4] - Quote
Your request for good gameplay is to reasonable. I demand less logic! MOAR TROLLING!!! |
Kitten Empress
Ametat Security Amarr Empire
392
|
Posted - 2013.06.09 06:43:00 -
[5] - Quote
Sontie wrote:Your request for good gameplay is to reasonable. I demand less logic! MOAR TROLLING!!! ARE YOU NOT ENTERTAINED?! THIS TOO?! |
Stefan Stahl
Seituoda Taskforce Command Caldari State
112
|
Posted - 2013.06.09 10:05:00 -
[6] - Quote
The abysmal netcode of Dust in comparison to Half-Life 1 is one of the main reasons I have been saying that CCP is about 10 years behind the rest of the industry.
People implement things like this in their spare time as a hobby. It's completely beyond me why I have to lead a target in Dust in 2013 to compensate for lag manually. |
Mobias Wyvern
Crux Special Tasks Group Gallente Federation
73
|
Posted - 2013.07.06 09:36:00 -
[7] - Quote
Be your own Judge. What game looks like more fun to you? Dust.........Or this.........
Planetside 2
Coming to PS4 this year
|
Imp Smash
Seraphim Auxiliaries CRONOS.
154
|
Posted - 2013.07.06 11:46:00 -
[8] - Quote
That kind of lag compensation leads to @curving@ bullets and dying behind corners or having your health fall from 70% to 0% in 0 seconds. It's damn annoying. if they could optimize the netcode so that more data is client side and less data is needed to be sent + is sent in smaller easier to use memory packets lag would drop significantly. THEN lag compensation like you describe would be more viable with fewer drawbacks. |
T3chnomanc3r
Ultramarine Corp
7
|
Posted - 2013.07.21 05:49:00 -
[9] - Quote
No, they need to kick people over a threshold OUT of the match or not let them in to begin with. The code is there to smooth occasional issues, not to make it a 250ms+ ping viable.
Imp Smash wrote:That kind of lag compensation leads to @curving@ bullets and dying behind corners or having your health fall from 70% to 0% in 0 seconds. It's damn annoying. if they could optimize the netcode so that more data is client side and less data is needed to be sent + is sent in smaller easier to use memory packets lag would drop significantly. THEN lag compensation like you describe would be more viable with fewer drawbacks.
|
T3chnomanc3r
Ultramarine Corp
13
|
Posted - 2013.07.23 08:11:00 -
[10] - Quote
Also since you put forth a terrible idea obviously without having done the required reading of the quoted article. We have enough people exploiting weakness of lag comp without adding more vectors:
Quote:The question arises, why is hit detection so complicated on the server? Doing the back tracking of player positions and dealing with precision errors while hit detection could be done client-side way easier and with pixel precision. The client would just tell the server with a "hit" message what player has been hit and where. We can't allow that simply because a game server can't trust the clients on such important decisions. Even if the client is "clean" and protected by Valve Anti-Cheat, the packets could be still modified on a 3rd machine while routed to the game server. These "cheat proxies" could inject "hit" messages into the network packet without being detected by VAC (a "man-in-the-middle" attack).
Imp Smash wrote:That kind of lag compensation leads to @curving@ bullets and dying behind corners or having your health fall from 70% to 0% in 0 seconds. It's damn annoying. if they could optimize the netcode so that more data is client side and less data is needed to be sent + is sent in smaller easier to use memory packets lag would drop significantly. THEN lag compensation like you describe would be more viable with fewer drawbacks.
|
|
Duck Drahko
DUST University Ivy League
65
|
Posted - 2013.07.23 11:02:00 -
[11] - Quote
I believe this emperor has no clothes...
What I mean is, I am quite certain that Dust includes lag compensation. I actually implemented this type of lag compensation in a few mods myself, back when it was still in theory mode, so I do believe that I would be able to tell if such an essential feature would be missing from Dust. So I'm not sure what you are all talking about.
The biggest network-related issue I have with Dust is that it doesn't show my current latency on the server (or even which region I am on, in case I am playing on Automatic or some random squad). I am forced to use a shared WIFI at the moment, so this would be quite important information. Also I get horrible latency on the US server, which makes the game hardly playable for me. But that is normal, no lag compensation can save you once your client view is sufficiently outdated and your reaction times suffer too much.
I haven't played a shooter in my life where playing on a different continent is not a severe disadvantage, no matter how good the lag compensation is. But without lag compensation, it would be entirely unplayable (which it is not). |
Imp Smash
Seraphim Auxiliaries CRONOS.
185
|
Posted - 2013.07.24 02:24:00 -
[12] - Quote
T3chnomanc3r wrote:Also since you put forth a terrible idea obviously without having done the required reading of the quoted article. We have enough people exploiting weakness of lag comp without adding more vectors: Quote:The question arises, why is hit detection so complicated on the server? Doing the back tracking of player positions and dealing with precision errors while hit detection could be done client-side way easier and with pixel precision. The client would just tell the server with a "hit" message what player has been hit and where. We can't allow that simply because a game server can't trust the clients on such important decisions. Even if the client is "clean" and protected by Valve Anti-Cheat, the packets could be still modified on a 3rd machine while routed to the game server. These "cheat proxies" could inject "hit" messages into the network packet without being detected by VAC (a "man-in-the-middle" attack). Imp Smash wrote:That kind of lag compensation leads to @curving@ bullets and dying behind corners or having your health fall from 70% to 0% in 0 seconds. It's damn annoying. if they could optimize the netcode so that more data is client side and less data is needed to be sent + is sent in smaller easier to use memory packets lag would drop significantly. THEN lag compensation like you describe would be more viable with fewer drawbacks.
I'm aware of how too much client side information can open the way for cheating.
Critical information should still be server side. I'm talking more a matter of data structure and lower priority information (such as your current position, actions (reload, hacking, ladders), etc.
Even without increasing client side load - data packets can be restructured to be smaller and more efficient. With the advances in memory that we had people generally take the easy way out with programming where if they took the time and effort they could send the same information with 40% less memory.
I'm referring to the structure more than the client / server load tbh. |
Duck Drahko
DUST University Ivy League
70
|
Posted - 2013.07.24 02:35:00 -
[13] - Quote
Imp Smash wrote:T3chnomanc3r wrote:Also since you put forth a terrible idea obviously without having done the required reading of the quoted article. We have enough people exploiting weakness of lag comp without adding more vectors: Quote:The question arises, why is hit detection so complicated on the server? Doing the back tracking of player positions and dealing with precision errors while hit detection could be done client-side way easier and with pixel precision. The client would just tell the server with a "hit" message what player has been hit and where. We can't allow that simply because a game server can't trust the clients on such important decisions. Even if the client is "clean" and protected by Valve Anti-Cheat, the packets could be still modified on a 3rd machine while routed to the game server. These "cheat proxies" could inject "hit" messages into the network packet without being detected by VAC (a "man-in-the-middle" attack). Imp Smash wrote:That kind of lag compensation leads to @curving@ bullets and dying behind corners or having your health fall from 70% to 0% in 0 seconds. It's damn annoying. if they could optimize the netcode so that more data is client side and less data is needed to be sent + is sent in smaller easier to use memory packets lag would drop significantly. THEN lag compensation like you describe would be more viable with fewer drawbacks. I'm aware of how too much client side information can open the way for cheating. Critical information should still be server side. I'm talking more a matter of data structure and lower priority information (such as your current position, actions (reload, hacking, ladders), etc. Even without increasing client side load - data packets can be restructured to be smaller and more efficient. With the advances in memory that we had people generally take the easy way out with programming where if they took the time and effort they could send the same information with 40% less memory. I'm referring to the structure more than the client / server load tbh.
No network programmer worth their salt is going to take the easy way out with packet sizes, especially not for a realtime shooter. And since Dust is based on the Unreal Engine, I think it's safe to assume that the network code is pretty darn optimised already. |
T3chnomanc3r
Ultramarine Corp
15
|
Posted - 2013.07.24 20:39:00 -
[14] - Quote
1. The client should show your as well as all other client's pings, major CPP mistake IMHO.
2. Lag comp in the form of rewind is clearly employed in Dust and IS giving the constant lagger an ADVANTAGE by letting them play in the past & changing already settled game events.
3, if your ping is bad, you don't belong in a game period. Lag comp like rewind & tolerance for pings >200ms should be sparing & fleeting AKA if it's not a few spikes out you go!
4. FW & PC are done through the London cluster, not regional so no manner of choosing region fixes that.
5. Instants are regional but that does not preclude a lagger from outside your region from playing and many do IMHO to game the lag comp system advantage.
6. All of this is about laggers gaining an advantage. Allowing pings above usable threshold allows CCP (like many console devs) to claim they have a huge base of players even if many of that base can't actually play each other.
FWI my ping (using traceroute) averages 160ms over 20 hops to 87.237.38.200, 140ms/13hops worth is Level3.net's backbone starting in Seattle on through to London. More than playable IMHO with the non-intrusive compression & client side prediction IF other players have similar <200ms connectivity. This is not an in game ping so YMMV but is sufficient for "mine is better" comparisons. :)
Duck Drahko wrote:I believe this emperor has no clothes...
What I mean is, I am quite certain that Dust includes lag compensation. I actually implemented this type of lag compensation in a few mods myself, back when it was still in theory mode, so I do believe that I would be able to tell if such an essential feature would be missing from Dust. So I'm not sure what you are all talking about.
The biggest network-related issue I have with Dust is that it doesn't show my current latency on the server (or even which region I am on, in case I am playing on Automatic or some random squad). I am forced to use a shared WIFI at the moment, so this would be quite important information. Also I get horrible latency on the US server, which makes the game hardly playable for me. But that is normal, no lag compensation can save you once your client view is sufficiently outdated and your reaction times suffer too much.
I haven't played a shooter in my life where playing on a different continent is not a severe disadvantage, no matter how good the lag compensation is. But without lag compensation, it would be entirely unplayable (which it is not).
|
Doc Noah
Algintal Core Gallente Federation
339
|
Posted - 2013.07.24 22:30:00 -
[15] - Quote
Dust's netcode is garbage, the person can be standing still and not register. Theres more than just 'lag' that is a problem. |
Duck Drahko
DUST University Ivy League
79
|
Posted - 2013.07.25 08:53:00 -
[16] - Quote
T3chnomanc3r wrote:1. The client should show your as well as all other client's pings, major CPP mistake IMHO.
2. Lag comp in the form of rewind is clearly employed in Dust and IS giving the constant lagger an ADVANTAGE by letting them play in the past & changing already settled game events.
3, if your ping is bad, you don't belong in a game period. Lag comp like rewind & tolerance for pings >200ms should be sparing & fleeting AKA if it's not a few spikes out you go!
4. FW & PC are done through the London cluster, not regional so no manner of choosing region fixes that.
5. Instants are regional but that does not preclude a lagger from outside your region from playing and many do IMHO to game the lag comp system advantage.
6. All of this is about laggers gaining an advantage. Allowing pings above usable threshold allows CCP (like many console devs) to claim they have a huge base of players even if many of that base can't actually play each other.
FWI my ping (using traceroute) averages 160ms over 20 hops to 87.237.38.200, 140ms/13hops worth is Level3.net's backbone starting in Seattle on through to London. More than playable IMHO with the non-intrusive compression & client side prediction IF other players have similar <200ms connectivity. This is not an in game ping so YMMV but is sufficient for "mine is better" comparisons. :)
Thank you for the information. I'm surprised that all competitive matches go through London (if that's really true), as that looks like a significant advantage for European teams (I'm not complaining since that's my region, but still...).
I used to get upset about high-pingers ruining the game, but not anymore. I guess I've just become more chilled over the years. ;)
I think the "rewrite history" concern isn't much of an issue, because the high-pinger has to deal with the same issue exaggerated. Sure it can be annoying for others, but not really unfair.
What still bugs me a bit is when high-pingers start warping around like crazy. That _is_ an advantage, because it makes it impossible to keep a rapid fire weapon on target without disruptions, and it's hard to line up a shot to begin with.
But to take advantage of that, one would have to combine a very fast suit with a weapon that requires little accuracy, maybe like a shotgun scout or mass driver scout. If I'd see somebody do this while consistently warping around, I would become suspicious... But all in all laggers still have it worse, and getting rid of them is not worth the trouble IMO. Not just would the game become much less inclusive, but it would also inevitably hit all of us occasionally, which would cause a lot of issues (we've probably all heard the jokes of the brother starting his **** download at the wrong time...). |
Himiko Kuronaga
SyNergy Gaming EoN.
1009
|
Posted - 2013.07.25 10:41:00 -
[17] - Quote
High ping players don't ruin games.
Games with bad netcode ruin games. If someone has a high ping it should be to their disadvantage and no one elses.
People need to take responsibility for their own equipment. |
T3chnomanc3r
Ultramarine Corp
15
|
Posted - 2013.07.25 11:50:00 -
[18] - Quote
Understand that all clients play their own copy of the world using ghosts & prediction based on the last word state update from the server server. Never forget that what you see is NOT always what is happening and world state history is revisionist for a window of "n" ms depending on lag comp code. When you see warp it's the result of data starvation beyond what your client's compensation can mask so the lagger disappears from your client.
The lagger client does NOT have same issue exaggerated they ARE THE ISSUE causing the effect as such see fluid action playing in the past against world events already unfolded then their actions change history. Only IF they lag beyond the compensation window or worse, interact with another lagged player, do they see the effects.
So warp is effect of server implicitly mandating to the clients to dump invalid predicted local world state forcing them to server world state. It's the rubberband effect when you are too lagged, and the warp effect when other clients briefly loose connectivity and/or lag.
Again, go read the Valve article before you speak further.
Duck Drahko wrote:I think the "rewrite history" concern isn't much of an issue, because the high-pinger has to deal with the same issue exaggerated. Sure it can be annoying for others, but not really unfair.
What still bugs me a bit is when high-pingers start warping around like crazy. That _is_ an advantage, because it makes it impossible to keep a rapid fire weapon on target without disruptions, and it's hard to line up a shot to begin with.
But to take advantage of that, one would have to combine a very fast suit with a weapon that requires little accuracy, maybe like a shotgun scout or mass driver scout. If I'd see somebody do this while consistently warping around, I would become suspicious... But all in all laggers still have it worse, and getting rid of them is not worth the trouble IMO. Not just would the game become much less inclusive, but it would also inevitably hit all of us occasionally, which would cause a lot of issues (we've probably all heard the jokes of the brother starting his **** download at the wrong time...).
|
T3chnomanc3r
Ultramarine Corp
15
|
Posted - 2013.07.25 12:19:00 -
[19] - Quote
It's likely not "their" equipment but rather a less than optimal route over the "best effort" Internet. As I've said, it's because CCP is allowing them to stay in game with an excessive average ping that ruins games. The lag comp is needed in small doses to smooth transient issues so it's application must be tightly controlled & used sparingly.
Himiko Kuronaga wrote:High ping players don't ruin games.
Games with bad netcode ruin games. If someone has a high ping it should be to their disadvantage and no one elses.
People need to take responsibility for their own equipment. |
T3chnomanc3r
Ultramarine Corp
15
|
Posted - 2013.07.25 12:39:00 -
[20] - Quote
LOL, you don't know if they're actually standing still because what you see in your client may not be the actual state.
Doc Noah wrote:Dust's netcode is garbage, the person can be standing still and not register. Theres more than just 'lag' that is a problem.
|
|
Duck Drahko
DUST University Ivy League
80
|
Posted - 2013.07.25 12:40:00 -
[21] - Quote
Alright, I should have known better than to try and have a reasonable discussion with a condescending prick. I am out now.
Just know that I am not a "console gamer", so your drivel is aimed at the wrong person. As a gamer I've been playing PC online shooters before lag compensation was a thing, and as a game developer I have implemented it in various engines when it did become a thing. Nobody knows everything, and I enjoy discussing these subjects with anybody, but the basic foundation is mutual respect, which you quite obviously lack.
Also, your top-postings are annoying. Have a nice day. |
Panther Alpha
DarkWingsss
768
|
Posted - 2013.07.25 12:48:00 -
[22] - Quote
Do you realize that lag compensation with the Source SDK is extremely hard to execute ? In counter strike only works to Improve the communication with the server, and reduce the packet lost.. but PING still is an issue regardless. Isn't really an effective way to lag compensate a FPS game.. It only works to improve the client / server communication to reduce packet lost.
The internet is going to have to get alot faster before lag can be compensated correctly. |
T3chnomanc3r
Ultramarine Corp
15
|
Posted - 2013.07.25 12:50:00 -
[23] - Quote
Sorry to hear ya go, truly! Sure I'm condescending but it's not drivel it's SCIENCE. I do so because what you repeatedly put forth does not reflect your claimed skills and I feel you ignore information when you post.
As to the top thing, sorry if that breaks your concept of presentation but we got to read your post above already I simply am quoting you with my response.
PS I too have been playing since Dom over IPX and Quake World first did TCP/IP over modems with no compensation.
Duck Drahko wrote:Alright, I should have known better than to try and have a reasonable discussion with a condescending prick. I am out now. Just know that I am not a "console gamer", so your drivel is aimed at the wrong person. As a gamer I've been playing PC online shooters before lag compensation was a thing, and as a game developer I have implemented it in various engines when it did become a thing. Nobody knows everything, and I enjoy discussing these subjects with anybody, but the basic foundation is mutual respect, which you quite obviously lack. Also, your top-postings are annoying. Have a nice day. |
Doc Noah
Algintal Core Gallente Federation
340
|
Posted - 2013.07.25 12:53:00 -
[24] - Quote
T3chnomanc3r wrote:LOL, you don't know if they're actually standing still because what you see in your client may not be the actual state. Doc Noah wrote:Dust's netcode is garbage, the person can be standing still and not register. Theres more than just 'lag' that is a problem.
Which proves my point, theres more than simply client1>server>client2 lag if the player position isnt even displayed correctly or the server 'loses' the player's position. In any other game, that would have either a)never have happened, b) kicked from server due to connectivity issues, or c) fixed in less than a month as its a huge problem in a FPS game. |
T3chnomanc3r
Ultramarine Corp
15
|
Posted - 2013.07.25 13:06:00 -
[25] - Quote
I agree with your there, that is what SHOULD happen IF a window of time passes but clearly that's a big window or non-existent here. Packet loss and cltent1>-server->client2 totally out of sync can & does happen without disconnect when the server is too tolerant.
Kicking is the fix IMHO.
Doc Noah wrote:T3chnomanc3r wrote:LOL, you don't know if they're actually standing still because what you see in your client may not be the actual state. Doc Noah wrote:Dust's netcode is garbage, the person can be standing still and not register. Theres more than just 'lag' that is a problem. Which proves my point, theres more than simply client1>server>client2 lag if the player position isnt even displayed correctly or the server 'loses' the player's position. In any other game, that would have either a)never have happened, b) kicked from server due to connectivity issues, or c) fixed in less than a month as its a huge problem in a FPS game.
|
|
|
|
Pages: 1 :: [one page] |