Hacker News Reader

Comment by incant | Hacker News Reader

About that 1800 players number. Based on posts by one of the developers, their server doesn't do any sort of calculation:

> Our server is basically 'dumb', it doesn't do anything in the sorts of processing bullet trajectory, player line of sight or any hit detection at all.

http://www.jc-mp.com/forums/index.php/topic,638.msg4783.html...

Unlike other games which run a full simulation on the server-side, their server (I presume) mostly routes messages between players. This means that the server trusts the client, which is rather poor design for a competitive multiplayer game. Once this is released, players will probably prefer a cheat-free game to a large-scale one.

This is, nevertheless, a fantastic mod. But these people have not yet solved the c10k-equivalent problem for massively multiplayer games.

View on HN

>which is rather poor design for a competitive multiplayer game

Watching that video, I would guess the environment isn't so much competitive (even single player JC2 is very no-particular-goals-in-mind-just-blow-shit-up oriented) as it is mass chaos. It looked like what would happen if /b/ bought an island. So cheating is probably only a minor concern at this point.

While cheating may not be a problem in a noncompetitive sandbox environment, griefing is. Consider a dumb client that fakes logging in to the game server and sends thousands of explosions to a player's screen at once, or drops hundreds of jets or explosive barrels everywhere instantaneously. There is practically no limit to sending these hand crafted packets in comparison to how fast the GPU will be able to render the resulting actions. So now, you have a massive DoS problem on your hands. Ideally nobody will do this, but my point is lots of problems can come up when you trust the client.

"drops hundreds of jets or explosive barrels everywhere instantaneously"

I don't know about you but this sounds like an awesome feature to me

The first time, sure. When it's happening every other minute it will just be dull.

DayZ is also not "competitive" at its core (and Arma2 even has anti-cheat measures built in -> BattleEye)... this doesnt stop script kiddies from server-hopping and killing/teleporting every one (even though they have no benefit from it beside the "lulz")..

Totally different game. DayZ is about solo play until you encounter other people, and even without cheating 90% of encounters with another person are shoot first, ask questions later.

Additionally, Just Cause is about just grabbing the nearest car and gun, then going out and causing chaos. DayZ requires players to spend literally days acquiring equipment, finding vehicles, fixing them. Finding a vehicle in DayZ can be a super rare find. In Just Cause 2, cars are everywhere.

Basically, cheating is a problem in DayZ because of the extreme scarcity of the resources. In Just Cause 2, there is no such scarcity. Therefore, cheating in DayZ ruins the game for everyone. Cheating in Just Cause 2 would probably go unnoticed, unless the cheater decided to become invincible.

They could verify those things on other clients, instead of on the server. I think that's typical for pure P2P multiplayer games. They could use something like - every player sending their version of 'what happened' to the server which then decides the proper outcome.

Btw, is it required to simulate everything on the server side? I know it is the conventional way to prevent cheating, but how about just randomly sampling some of the actions to detect cheating, and ban that user?

"They could verify those things on other clients..."

But clients are not trustworthy. How do you decide which clients to trust? How long before your 'adversary' discovers your trust methods an replicates them in his client?

"...server which then decides the proper outcome."

This requires processing power. Random sampling is perhaps a possible solution, but if you're overwhelmed with nefarious client activity (50% perhaps) and you're only sampling, say, 10%, still 40% of your gameplay is cheating clients. If you increase your sampling based on finding increases in cheating, you start to overwhelm your server farm.

tl;dr - this is not an easy problem to solve

If 50% of your clients are cheating, are you really playing a game?

If you can write a little C program that speaks enough of the game's wire protocol to simulate a bot, send nasty packets, and also provide a vote of confidence when other bots ask if those packets are correct--and then run 100000 of them over a few machines--then 99.999% of your clients could be "cheating" while there are still people trying to play a real game.

One major problem with this is there are plenty of legitimate communications that will originate from a single client -- namely, those client's in-game actions!

Any cheating that takes the form of something "possible" for a client to do would be undetectable under such a scheme. (eg aimbotting)

On the other hand it's not like usual FPS have everything calculated on the server side, which is why there is a difference between most MMOs and FPSs and why there are still cheaters. There are various countermeasures. There are even a lot of MMOs that do a big part on the client side.

Maybe clients could also verify each other. There has been some thought on this topic, but I don't how far the research went.

Can't the server mitigate a lot of this cheating by enforcing a matching checksum? That's how Subspace (AKA Continuum) does it, and cheating seems to be fairly minimal.

rsanchez1 and jjrumi both posted reasonable answers, but are hellbanned.

Guys: You need to email an admin and get yourself unbanned. Most people don't see your posts.

Reply on news.ycombinator.com