Sniper behavior in VC:MP (and more)

[karan]

VSYNC needs to be improved, fps drops when it comes to custom objects, could've agreed if it's because of wooden pc but I feel the same when playing from my laptop which is having sufficient graphics to run battlefield at semi maxed settings but fail to run vcmp smoothly, whereas I remember playing this game smoothly even with my P4 machine back in 2011, but for now, it requires more than sufficient GPU usage and I guess you can do something about it.

Ping issues are there, F5 shows fake ping tbh, I've played 0.3 a few days back, my ping and rest pings were 200+ on F5 scoreboard, whereas when I joined 0.4 server, it shown somewhat 170-180 ping, dunno what is it, but 0.4 shows low ping than 0.3z used to show. If a guy is pinging 250 in 0.3z server, he pings 200 in 0.4, maybe update rate in 0.4 is decreased.

[AsaD.]

You can kill a high ping player more easily than killing a guy whose FPS is low. I agree with Charley, we can see the anims of a high ping player better than low FPS player. Killing a high ping player is difficult but killing a low FPS player is impossible(unhitabble).

nah thats not right.

[AsaD.]

The Sliding is a big issue,when laggers slide then its really unable to kill them.So my suggesion is to disable slide first to make sure that killing laggers is possible.

[KakaroT]

I meant my whole VC:MP career(not sure if this is the right word).But thats okay since you arent capable enough to understand this.

[krystianoo]

VSYNC needs to be improved, fps drops when it comes to custom objects, could've agreed if it's because of wooden pc but I feel the same when playing from my laptop which is having sufficient graphics to run battlefield at semi maxed settings but fail to run vcmp smoothly, whereas I remember playing this game smoothly even with my P4 machine back in 2011, but for now, it requires more than sufficient GPU usage and I guess you can do something about it.

Ping issues are there, F5 shows fake ping tbh, I've played 0.3 a few days back, my ping and rest pings were 200+ on F5 scoreboard, whereas when I joined 0.4 server, it shown somewhat 170-180 ping, dunno what is it, but 0.4 shows low ping than 0.3z used to show. If a guy is pinging 250 in 0.3z server, he pings 200 in 0.4, maybe update rate in 0.4 is decreased.

Custom objects can lag if there are a lot of them and they use autocollision type instead of manual collision. Autocollision is excellent collision-wise but a lot of objects using it make the game work really poor performance-wise.

[JuaN.]

ghost shot

[Sora_Blue]

VSYNC needs to be improved

Who the fuck uses VSync in 2017?

Ping issues are there, F5 shows fake ping tbh, I've played 0.3 a few days back, my ping and rest pings were 200+ on F5 scoreboard, whereas when I joined 0.4 server, it shown somewhat 170-180 ping, dunno what is it, but 0.4 shows low ping than 0.3z used to show. If a guy is pinging 250 in 0.3z server, he pings 200 in 0.4, maybe update rate in 0.4 is decreased.

No this has been explained before, the ping values in 0.4 are actually more accurate compared to 0.3, from what I can recall Storm said that 0.3 used an estimate or some shit.

[karan]

VSYNC needs to be improved

Who the fuck uses VSync in 2017?

Well, I need to turn vsync on from gpu settings and frame limiter off from vc settings, if i turn off frame limiter only and leave vsync off, my game goes nuts, and then custom objects play a great role in making it more terrible, there should be a solution to this or is it happening because of some particles?

No this has been explained before, the ping values in 0.4 are actually more accurate compared to 0.3, from what I can recall Storm said that 0.3 used an estimate or some shit.

lol nope, even [EAF]King was surprised the day he joined vcmp 0.4 for the first time, because he pings 350+ everywhere but 0.4 made his ping somewhat 270, this accuracy isn't accurate but slow update rate.

[stormeus]

Am I correct in thinking this is perhaps theoretically an impossible issue to solve? Is it the case that low-FPSers will be sending fewer anim packets thereby skipping some, and that jittery-pingers might be sending them, but there is no way that they can be received? If I'm wrong and there is some sort of solution, then that would be great.

VC:MP actually doesn't sync animations individually. It syncs what keys you're pressing at the moment the sync packet gets sent, and then other clients use that and let the game process the animation sequences based on those keystrokes.

The problem with people with low framerates is that because Vice City processes change based on how many frames have elapsed as opposed to how much time has passed since the last frame, they miss crucial information like when a player has just fired their weapon. This is fundamentally unfixable due to how the sync and frame processing work.

The problem with jittery pings is that if one packet suddenly takes 1500ms to reach the server, and then the next sync packet takes 300ms, the second packet will reach people before the first. Because sync packets are unordered, this might cause momentary out-of-order animation processing that causes sliding. I'm not concerned about this scenario, and while we can make sync packets ordered this can also cause more issues than it fixes by causing processing delays when dealing with laggy players, since the game would have to wait for the next packet in the sequence.

edit
I should also mention that the problem with trying to sync the animations themselves is that in the time between sync packets someone's animations may have already changed too much for other players to pick up on. When testing animation progress sync on localhost, SMGs and machine guns almost completely broke because of how the game processes those animations. This is somewhat of an oversimplification.

No this has been explained before, the ping values in 0.4 are actually more accurate compared to 0.3, from what I can recall Storm said that 0.3 used an estimate or some shit.

lol nope, even [EAF]King was surprised the day he joined vcmp 0.4 for the first time, because he pings 350+ everywhere but 0.4 made his ping somewhat 270, this accuracy isn't accurate but slow update rate.

I don't remember this conversation. 0.4 uses the exact same method of measuring ping as 0.3, and in both cases they're handled by the network layer, not us.

[aXXo]

I don't remember this conversation. 0.4 uses the exact same method of measuring ping as 0.3, and in both cases they're handled by the network layer, not us.

The ping shown in 0.3z was wrong. (VCMP browser showed correct, ingame F5 was wrong)
I remember it showed 30ms even on localhost, when it should be 0.

I think the ingame display added a certain amount to everyone's ping depending on number of hops to the server. No-one used to have less than 10ms ping in 0.3z while we have spotted some players with less than 5ms ping in 0.4. Thijn or Brian could confirm.

[karan]

I don't remember this conversation. 0.4 uses the exact same method of measuring ping as 0.3, and in both cases they're handled by the network layer, not us.

The ping shown in 0.3z was wrong. (VCMP browser showed correct, ingame F5 was wrong)
I remember it showed 30ms even on localhost, when it should be 0.

I think the ingame display added a certain amount to everyone's ping depending on number of hops to the server. No-one used to have less than 10ms ping in 0.3z while we have spotted some players with less than 5ms ping in 0.4. Thijn or Brian could confirm.

Opposite scene in 0.4, sometimes it show 350+ ping on browser but when join ingame, it become 160 on F5 ._.

[krystianoo]

Once you patch Snipers to shoot only through the glass, make sure that when there is a glass ----> wall, snipers won't penetrate the glass and the wall afterwards.

[Sevrin]

I don't remember this conversation. 0.4 uses the exact same method of measuring ping as 0.3, and in both cases they're handled by the network layer, not us.

The ping shown in 0.3z was wrong. (VCMP browser showed correct, ingame F5 was wrong)
I remember it showed 30ms even on localhost, when it should be 0.

I think the ingame display added a certain amount to everyone's ping depending on number of hops to the server. No-one used to have less than 10ms ping in 0.3z while we have spotted some players with less than 5ms ping in 0.4. Thijn or Brian could confirm.

Kyzo, Brian, Murdock, Thijn, Salmonella and many guyd who play in Argo have ping less than 10ms.

[stormeus]

I don't remember this conversation. 0.4 uses the exact same method of measuring ping as 0.3, and in both cases they're handled by the network layer, not us.

The ping shown in 0.3z was wrong. (VCMP browser showed correct, ingame F5 was wrong)
I remember it showed 30ms even on localhost, when it should be 0.

I think the ingame display added a certain amount to everyone's ping depending on number of hops to the server. No-one used to have less than 10ms ping in 0.3z while we have spotted some players with less than 5ms ping in 0.4. Thijn or Brian could confirm.

GetPlayerPing and the equivalent Squirrel function should've been accurate. I'm​ sure the scoreboard display had issues, though, considering the code from 0.1 to 0.3 looked like this:

https://bitbucket.org/stormeus/vice-players/src/7c74ebe935fd5934ae2143150f9b1f60dd075d2a/server/netrpc.cpp?at=master&fileviewer=file-view-default#netrpc.cpp-304

I'm fairly confident they either mixed up player index and system address at some point or they had an off by one error. If I recall correctly, the scoreboard was also prone to mixing up people's pings.

The whole thing was messed up and I'm surprised anyone thought it'd be a good idea to send everyone's IP addresses to everyone else too. It's messy architecture all the way down.

Kyzo, Brian, Murdock, Thijn, Salmonella and many guyd who play in Argo have ping less than 10ms.

The server is in the Netherlands, Thijn is in the Netherlands. Big shock 

[Milko]

that signature...