Anti-cheat

[Sniper]

For the next version how about some limitations with how high/low you can set things. I know its not exactly punkbuster but I mean there is a fine line between settings your comfortable with and basically cheating.

[Belier13]

I have never seen any cheater.

[Howitzer]

At least mention what variables
Should rate be limited?, should cg_maxfps be limited?, ...

[Sniper]

Heck no! I love 43 fps:D But pmove_fixed should be made 0. err...There is an extreme picture somewhere but I cant find it it has gamma 5 foliege 0 and stuff

[vcxzet]

most of the variables have upper and lower limits
setting something higher/lower does not mean anything at all.
well a skilled(!) programmer can modify the exe for custom views
player/building location markers,aimbots,autoshooters ie everything that is client side.

[Paradox]

You mean that some player took the time to build skills, the responses, and the knowledge to set their mouse's sensitivity so high that by blowing on it they do a 360 is cheating! Well then, driving must be cheating, because only people above 16 can do it!

[David]

any kind of client side anti-cheat would have to be closed source, otherwise it would be 'cracked' very quickly.

[vcxzet]

any kind of client side anti-cheat would have to be closed source, otherwise it would be 'cracked' very quickly.

lol if someone wants it so badly you cant prevent it. anything is opensource in asm

[kevlarman]

any kind of client side anti-cheat would have to be closed source, otherwise it would be 'cracked' very quickly.

lol if someone wants it so badly you cant prevent it. anything is opensource in asm

not really, it may be possible to disassemble the binary, but gaining anything useful from it is hard, especially when gcc gets to do creative optimizations, and if you really don't want people disassembling your code, you can encrypt it in the file, and have it decrypted only once it's in memory (though a memory dump will get around this). There are ways to make an open source anti cheat be difficult to crack, but they would violate the 'keys that are required to make a program work must be provided with the source' clause of the GPLv3 and/or include binary blobs.

[vcxzet]

any kind of client side anti-cheat would have to be closed source, otherwise it would be 'cracked' very quickly.

lol if someone wants it so badly you cant prevent it. anything is opensource in asm

not really, it may be possible to disassemble the binary, but gaining anything useful from it is hard, especially when gcc gets to do creative optimizations, and if you really don't want people disassembling your code, you can encrypt it in the file, and have it decrypted only once it's in memory (though a memory dump will get around this). There are ways to make an open source anti cheat be difficult to crack, but they would violate the 'keys that are required to make a program work must be provided with the source' clause of the GPLv3 and/or include binary blobs.

lol they even crack 10000$ very well protected softwares in days. 99% protection methods were proved useless. and people are working on the rest %1

[kevlarman]

you're right, any anti-cheat system can be broken with enough effort, but the thing is, who is going to put that much effort into tremulous?

[vcxzet]

you're right, any anti-cheat system can be broken with enough effort, but the thing is, who is going to put that much effort into tremulous?

I know many people with too much knowledge and time. Fortunately they are not playing trem

[rasz_pl]

we dont need client side anti cheat, server side is enough. Remember =- server sees EVERYTHING you do, all it takes is one "unusual" action and bam, server can autokick.

[Caveman]

lol, why do you think PB went client-side?
But anyways _every_ cheatprotection is at a loss when allowing clients that run on a selfmade OS .)
As long as a client has total control over what is sent to the server cheatprotection is laughable.

Besides that Trem has allready all that is needed to make it very hard for cheaters.
- Speedwalk? for a q3 engine? lol
- Wallhack? Certainly possible as it is already available and impossible to fight due to the way such a cheat can be hooked. But so what? it would only help humans in S1...
- Aimbot? Who wants to go to that extreme and do a complete recoding for movement prediction? Remember we only have one weapon that has insta-hit the rest is delayed. And an aimbot for alien? Come on, get real

Summing this up I really can't see neither a need nor a way to protect trem from cheats.

[kevlarman]

lol, why do you think PB went client-side?
But anyways _every_ cheatprotection is at a loss when allowing clients that run on a selfmade OS .)
As long as a client has total control over what is sent to the server cheatprotection is laughable.

Besides that Trem has allready all that is needed to make it very hard for cheaters.
- Speedwalk? for a q3 engine? lol
- Wallhack? Certainly possible as it is already available and impossible to fight due to the way such a cheat can be hooked. But so what? it would only help humans in S1...
- Aimbot? Who wants to go to that extreme and do a complete recoding for movement prediction? Remember we only have one weapon that has insta-hit the rest is delayed. And an aimbot for alien? Come on, get real

Summing this up I really can't see neither a need nor a way to protect trem from cheats.

an aimbot could be a reasonably useful cheat (lasgun/md/shotgun, and you get s2 very fast), but you do bring up a good point, and because of the problem of griefers, tremulous has many more admins than most games, so any cheating will get you a prompt ban from that server.

[PFB]

HPUN!

[a Turret]

HPUN!

Tremulous will never use Punkbuster, Costs to much and Tremulous is open source and then we would be able to see the PuinkBuster code which could be used to cheat in their other games.

[TinMan]

I've only seen aimbots and wallhacks, but neither really help you that much in trem. Aimbots usually don't hit the target well and wallhacks would just be confusing to use opposed to radar.