rbrb
| Crates.io | rbrb |
| lib.rs | rbrb |
| version | 0.0.1 |
| source | src |
| created_at | 2021-11-23 16:05:16.666044 |
| updated_at | 2021-11-23 16:05:16.666044 |
| description | Write robust rollback networked games |
| homepage | |
| repository | https://github.com/shelbyd/rbrb |
| max_upload_size | |
| id | 486317 |
| size | 3,124 |
Shelby Doolittle (shelbyd)
documentation
README
rbrb
A library for building RoBust RollBack-based networked games.
rbrb is heavily inspired by GGPO and
GGRS, but aims to be more reliable and capable.
Assumptions
This library assumes your game is a deterministic Fn(&State, Set<Input>) -> State.
We (will) have an additional testing mode that will spend extra cycles on checking that the
state is consitent between players and deterministic on the same logical update.
Roadmap
Core Functionality
- Multi-party sync
- Consistent disconnection
- Reconnect disconnected player
Robustness
- Determinism checks
- Checksum propagation
- Debugging failed checks
- Fake a bad network
- Confirmation state
Features
- In-game replays
- Out of game replays
- Headless
- Spectators
- Drop in/out
Performance
- Input delta encoding
- Hub and spoke network
License: MIT