

#1
Posted 26 May 2025 - 04:45 PM
This is not the first time I have heard of this or seen this. I know it has been brought to PGIs attention. Now fix it.
#2
Posted 26 May 2025 - 05:34 PM
It's velocity being slightly delayed reflection on your screen vs what the server thinks it's velocity, and thus position is, coupled with your travel time and spread weapon, will likely result in you not hitting very much.
Stop doing that.
#3
Posted 26 May 2025 - 06:31 PM
#4
Posted 26 May 2025 - 10:18 PM
Edited by Saved By The Bell, 26 May 2025 - 10:19 PM.
#5
Posted 26 May 2025 - 10:19 PM
I mean hey, I managed to hit a jumping Spider at 700m once with a gauss, so it's not like it's entirely impossible to hit.
Edited by Ttly, 26 May 2025 - 10:21 PM.
#6
Posted 26 May 2025 - 10:29 PM
#7
Posted 26 May 2025 - 10:31 PM
I mean, that's horrible for a monitor response rate, but I really don't know about server rates.
#8
Posted 27 May 2025 - 01:17 AM
Void Angel, on 26 May 2025 - 10:31 PM, said:
I mean, that's horrible for a monitor response rate, but I really don't know about server rates.
It means it exchanges data with any connected clients 30 times per second and updates the game state at the same rate. This is why there is a plethora of predictive code and as long as mech speed and weapon velocity stay within a certain range its fine. However power creep on velocity and speed over the years is pushing this to the limits. Plus hitreg on twisting jj mechs has always been dodgy hence why you see it done even now.
#9
Posted 27 May 2025 - 01:42 AM
Meep Meep, on 27 May 2025 - 01:17 AM, said:
It means it exchanges data with any connected clients 30 times per second and updates the game state at the same rate. This is why there is a plethora of predictive code and as long as mech speed and weapon velocity stay within a certain range its fine. However power creep on velocity and speed over the years is pushing this to the limits. Plus hitreg on twisting jj mechs has always been dodgy hence why you see it done even now.
I know what it means. =] I just don't know if 30hz was state of the industry at the time, and how it compares now.
#10
Posted 27 May 2025 - 02:32 AM
#11
Posted 27 May 2025 - 05:10 AM
GreyNovember, on 26 May 2025 - 05:34 PM, said:
It's velocity being slightly delayed reflection on your screen vs what the server thinks it's velocity, and thus position is, coupled with your travel time and spread weapon, will likely result in you not hitting very much.
Stop doing that.
Well, you could advise him on how to compensate for that? If the enemy mech's position on screen isn't accurate to where it actually is, any ideas on how to help the guy improve?
"stop doing that" is not helpful......
#12
Posted 27 May 2025 - 06:20 AM
Kamiko Kross, on 27 May 2025 - 05:10 AM, said:
"stop doing that" is not helpful......
The advice is to not shoot when you're unlikely to hit + And then raise a stink because your unlikely shot failed. Don't go to a stranger's BBQ expecting to be accommodated as a vegetarian.
The mech is in a temporary state of high evasion. Wait for it to eventually drop from the sky before you try to hit it with a weapon that is inherently inaccurate, while riding a mech that is presumably also running and twisting to keep track of the spider.
It'd be best if he also did not try to engage the spider after running directly into a crowd of 3-4 mechs waiting in a corner of Grim Plexus that were supporting it, staying in range of those mechs, and getting Lurmed to death in his Viper. The lack of stress from that situation would greatly assist and reduce the factors that contribute to failing to hit the target.
Edited by GreyNovember, 27 May 2025 - 06:54 AM.
#13
Posted 27 May 2025 - 07:58 AM
Meep Meep, on 27 May 2025 - 01:17 AM, said:
You're conflating some things. MW4 had way faster velocities on average (2000m/s was the slowest, with up to 3750m/s vanilla) and hit detection worked fine as long as you knew how to lag shoot. It just matters on collision detection techniques and ensuring that client and server are synced as much as possible to ensure responsiveness. This is why CS2 mostly stayed 64 instead of 128 because they added subtick which is very similar to HSR (in that it is latency compensation). Latency compensation will never be perfect which is why you see some deviances in it (because it's trying to hit a moving target effectively), and slower/smaller projectiles will feel it the most because the issue isn't hit detection, it's netcode related. i've noticed myself having to lag shoot with lasers because HSR isn't compensating well enough in one direction (either it's overcompensating for my latency or not compensating enough).
tl;dr server/client netcode is a distributed system and that stuff is hard to do "right" and is likely to blame for a lot of the wonkiness of hit registration in this game. That and interpolation, interpolation can also be fun depending on the amount of time you are trying to interpolate
Edited by Quicksilver Aberration, 27 May 2025 - 08:03 AM.
#14
Posted 27 May 2025 - 11:14 AM
Quicksilver Aberration, on 27 May 2025 - 07:58 AM, said:
This was a lovely thing to try and figure out when making a system that used preset animations, in a game where people could cancel out of said animations and rapidly reposition, in a P2P environment, based in a third world country where your internet connection could be at best described as "spotty".
How far do we look ahead into the future, how often do we update, and how drastic of a change can the variables make between checks?
In a perfect world we'd be brute forcing it with infinite power and almost zero latency to "Fix" it, as is commonly said.
#15
Posted 27 May 2025 - 12:03 PM
Quicksilver Aberration, on 27 May 2025 - 07:58 AM, said:
tl;dr server/client netcode is a distributed system and that stuff is hard to do "right" and is likely to blame for a lot of the wonkiness of hit registration in this game. That and interpolation, interpolation can also be fun depending on the amount of time you are trying to interpolate
But do the numbers they show us actually relate to their real in game velocities as rendered in the two games? Mw4 was on an entirely different engine(MUNGA) and not this cryengine spaghetti monster of hacked code and bandaids to wrestle it to do what they wanted it to do. Hitreg has been an issue since beta and they knew even back then that too much velocity or speed would start to break things.
#16
Posted 27 May 2025 - 01:04 PM
#17
Posted 27 May 2025 - 01:08 PM
Meep Meep, on 27 May 2025 - 12:03 PM, said:
But do the numbers they show us actually relate to their real in game velocities as rendered in the two games? Mw4 was on an entirely different engine(MUNGA) and not this cryengine spaghetti monster of hacked code and bandaids to wrestle it to do what they wanted it to do. Hitreg has been an issue since beta and they knew even back then that too much velocity or speed would start to break things.
youve never been under the hood have you? all game engines are full of spaghetti code, its the nature of the beast. i remember when they opened the freespace engine, it was full of gotos. every one of them was a necessary optimization to get out of a loop in a hurry to do something critical elsewhere. single threaded engines were fun.
#18
Posted 27 May 2025 - 05:22 PM
LordNothing, on 27 May 2025 - 01:08 PM, said:
FTFY
Meep Meep, on 27 May 2025 - 12:03 PM, said:
HSR allowed poptarts to be a thing because you no longer needed to lag shoot significantly to reliably use them vs lasers. Velocity can be an issue depending on the interpolation of the hit detection but you would also see that issue in training grounds and academy (if you cap FPS to 30 that is). I'd hazard a guess the issue is just netcode and having to deal with state (like client/server desync).
Edited by Quicksilver Aberration, 27 May 2025 - 05:44 PM.
#19
Posted 27 May 2025 - 10:03 PM
LordNothing, on 27 May 2025 - 01:08 PM, said:
Just saying that this game has an extra cheese layer of spaghetti and the cryengine base they used wasn't even fully patched. The devs did wonders though and the game mostly works like it should. Mostly..

#20
Posted 27 May 2025 - 10:32 PM
1 user(s) are reading this topic
0 members, 1 guests, 0 anonymous users