Hacker News new | ask | show | jobs
by quacker 3025 days ago
I have a few reactions here:

1. Why should Epic go and reimplement anti-cheat that you can grab off the shelf?

2. A big part of limiting cheating comes down to the specific game design and implementation. You need a server implementation that is as distrusting of clients as possible, you need secure (networking) code, etc. Anti-cheat is not a cure-all.

3. Even without anti-cheat, Unreal Engine is still well-used by tons of single-player video games. I don't see it "unraveling" regardless.

1 comments

> 2. A big part of limiting cheating comes down to the specific game design and implementation. You need a server implementation that is as distrusting of clients as possible, you need secure (networking) code, etc. Anti-cheat is not a cure-all.

Unfortunately that's just not realistic for a large category of games. If you can trust the clients to some degree then you can really really offload a lot of server work. It's not feasible to process user's mouse input on the server in an FPS for example. Not just from a server load standpoint but from a latency standpoint.

You don't need to do this in real time, instead you could log the data (e.g. mouse clicks, key presses) to something like a database, Kafka, or Kenesis. Now you've unloaded the anti cheat logic to other servers.

You don't necessarily need to ban a player immediately. If you can accurately ban people within a day of using a cheat, then that's a pretty serious deterrent from using them.