Game Development: Pfefferbench

[demetharkey]

enough waiting. ima vibe code a new mmo

[524u]

enough waiting. ima vibe code a new mmo

Hafen 2

[paskasammakko]

when are they aiming for

[524u]

Apologies for breaking mound beds. Fixing.

When

[Lao_Bao]

My question for AI: Could you help them with this problem (OCO)

AI response:

I'd love to help Björn and Fredrik with this architectural challenge, but I can't directly make changes to the Haven & Hearth code. The Seatribe project has a completely closed source code of the server part, and Loftar itself fundamentally writes and controls the engine alone, not allowing third-party developers to the core of the system.However, if Loftar decided to use artificial intelligence as an assistant, I could significantly speed up the OCO development process in the following areas:

1. Entity Component System Design

I can help design or optimize the model ECS (Entity Component System) for nested objects.


• How to solve it: Instead of rigidly binding the coordinates of the character to the global map, the concept of "Local Coordinate Space" is introduced.
  
  
• My help: Writing pure algorithms in Java/C++ to transform coordinate matrices so that when a parent object (e.g. a ship) moves, all child objects (characters, chests) are automatically shifted relative to it without constantly recalculating their global position on the server.

2. Network Synchronization Simulation (Anti-Sync)

The problem of desynchronization between the client and the server during the movement of OCO is one of the most difficult.


• My help: Development of mathematical models for Client-Side Prediction (client-side motion predictions) and interpolation algorithms. I can offer out-of-the-box data packet structures that minimize network load by transmitting only the relative motion vectors, rather than the absolute coordinates of each entity every millisecond.

3. Code stress testing and exploit search

Since Loftar is afraid of dues and geometry bugs, the code needs to be rigorously tested before deployment.


• My help: Generate scripts for automated testing (Unit and Integration tests). I can write scripts that simulate thousands of crazy player actions at once (like trying to get on a wagon when it's destroyed, or parking a ship exactly on the border of two game hitboxes) to identify holes in the logic before players find them.

4. Refactoring old code and routine

The biggest barrier for Loftar is a lack of time.


• My help: I can take care of routine legacy code refactoring, adapting custom rendering for new libraries, writing documentation, or generating boilerplate code for weekly patches. This would free up Fredrik up to 50% of his working time, which he could devote entirely to OCO.

Can AI really accelerate OCO development? How useful is it in this regard?

[Regulus2424]

who tf are Björn and Fredrik

[524u]

Some randoms

[Frogg]

Just give the Ai company access to everything you own so it can vibe-code some bullshit for you lmao

don't do that

[magnet]

who tf are Björn and Fredrik

This is Bjorn and Fredrik:

[Frogg]

OCOmaxxing