There’s so many cheaps that are only possible because the game’s net code is just weirdly implemented.
Like wall hacks, why can I see this other player on the other side of the map, why you sending me their position.
Comment on Kernel Anti-Cheat Is an Overreach
warm@kbin.earth 3 weeks ago
Anti-cheat should be server side. Cheats are done on external hardware now as kernel AC chased them into the undetectable zone, well played.
No game is worth sacrificing your entire PC to play.
There’s so many cheaps that are only possible because the game’s net code is just weirdly implemented.
Like wall hacks, why can I see this other player on the other side of the map, why you sending me their position.
it’s for speed. if I tell you they’re out of sight now, but will be visible in 50ms, it’ll be faster than waiting 50ms to recv their new position that’s visible and waiting on network latency which could be another 50ms in addition
I’ve been working on a PvP game for the last few months.
Visibility checks aren’t needed in all cases. Most RTS/2D/2.5D games don’t have this problem, or it is minimal.
For games that do need them, they tend to be AAA FPS, which have the budget to make simple changes like lag compensation.
Right but we’re talking about 100s of ms of human reaction time (faster reaction time sis around 200ms for ultra pros). Wall hacks where you only get 400ms of warning aren’t much good. So they absolutely could do something with it.
100ms more is absolutely enough to be massive because it just subtracts from however long you need to recognize the other player after they become visible. Not to mention that in slower paced shooters, people might just remain stationary near a corner for very long, and you can’t magically know when they start to move. Also in games where you can die with one well placed shot it could be 10ms and still be a significant advantage because you just need to be faster than your opponent.
Afaik valorant does try to not send any info that won’t be needed, that doesn’t mean it’s immune to wallhacks, it just limits their effectiveness.
They can see each other though, which means one player may get information about the enemy 100ms ahead of another. That would be quite an advantage in a lot of games, especially in Rainbow 6 Siege where TTK is super short.
Cool I’m an ULTRA PRO and i’ve been drinking for a good chunk of today
Culling is already done to reduce wall hacks, but a) sound needs an origin and b) people notice the difference between 3 frames and 6 frames reaction (1f =16.6ms)
Same reason I don’t use the Hypervisor Denuvo bypasses.
thingsiplay@lemmy.ml 3 weeks ago
Server side anticheat can’t detect lot of client side anticheat. Ideally you want to have both, if you want to catch most cheaters. One is not a full replacement for the other method. Local anticheat has the advantage of being much cheaper for the developer / publisher.
Sir_Thominick_IV@lemmy.world 2 weeks ago
Depending on the implementation, client side anticheat isn’t needed at all. Take the game Speedrunners for example. There is zero need for the server to do anything other than accept player inputs, process them, and return them to the clients.
So if coded correctly, where the server trusts nothing and does the math itself, it is impossible to cheat while having zero anticheat on the client.
GoatSynagogue@lemmy.world 2 weeks ago
That’s all well and good, but a smash brothers clone is very different to CoD or Battlefield in this regard.
Client side anti-cheat is absolutely needed in FPS/TPS. Without it, server side anti-cheat would have to do far more and quietly be much harder to do.
Sir_Thominick_IV@lemmy.world 2 weeks ago
Those types of games are usually made by AAA studios. They have the resources to do server side anticheat despite it being more involved.
warm@kbin.earth 3 weeks ago
It's more costly to the developer when I dont buy their kernel AC game.
thingsiplay@lemmy.ml 3 weeks ago
Unfortunately we are in the minority. Most people don’t know the difference or even care. So like with physical media, they only care about the majority.
warm@kbin.earth 2 weeks ago
Sure, in the grand scheme of things, but per user, not getting $70 from me costs them more.