but they dont know which flag is being used to check for anticheat because the flag checking/sending is being handled inside the closed anticheat files
client and server both would have a closed side for anticheat, and the flags for anticheat would be inside that closed side only and flags are sent with encryption
I'd prefer to have teehistorian and some anti cheat models(ml) that can be passed around, which would not rely on any closed source stuff. It would allow to crowd source bot input data and also allow ML people to work with that data in order to create better detection models. I kind of doubt such models can be easily reverse engineered, as there might simply be too many variables to do so(not a ml person). :( give me the open source variant!
16:50
the trained model files could be "closed source"(edited)
Usually cheat creators operate as individuals, currently noby also operates as an individual. if we have another 0.6 bot endgame, i'd guess it's too much work for one individual to "fight" against multiple small teams or individuals, thus the crowdsourcing idea. whole community providing data (should be confirmed by people in any case) in a fight against individuals creating cheats. if there would be an interchangable default way for such data to be exchanged, this would imo help.
There is a bunch of 0.6 bots openly available, that people use. it's not like everyone still playing on 0.6 creates their own bot to bypass that detection.(edited)
17:17
some people might be not that bright to attempt to use the same bot again, even tho they were banned previously.
17:19
laying out the groundwork to prevent a 0.6 endgame, might help on the long run with 0.7(edited)
[11:59 AM] Kaisi: even dyncam players
wtf i play with dyncam what
20:24
and yes there is no such thing as a perfect antibot
20:25
everything can be bypassed in theory
20:25
its just a fun project and at the very least reduces how powerful the bots can be
20:26
and the arms race is always gonna be a thing even if there is no automated system
20:26
replace "server side antibot" with "admin watching the server" and the same idea could apply where people try to make better bots that the admin wont see
I am working on an open source C++ project, for code that compiles on Linux and Windows. I use CMake to build the code on Linux. For ease of development setup and political reasons, I must stick to
when do we get anticheat?
