We have been using Fyrox for a community project at work for over a year now. We run into bugs from time to time, but for the most part it works if you are willing to put in a bit more effort.

[–] entwine@programming.dev 1 points 1 day ago (2 children)

Why pick Fyrox? Not to shit on rust, it clearly has its place, but looking at the track record of game engines, video games aint it.

[–] SorteKanin@feddit.dk 1 points 14 hours ago (2 children)

Do you mean Rust is unsuited for games, or that Rust simply don't (yet) have the established engines/frameworks/libraries for it? I think Rust could be really great for video game development, but it's true that the ecosystem is not quite there yet - though it's getting better.

[–] entwine@programming.dev 3 points 8 hours ago (2 children)

I believe Rust is a poor choice for games. Games have always needed to balance iteration times with performance. C++ has traditionally been a good middle ground, but nowadays computers/consoles are fast enough that even javascript is a solid choice.

Rust makes sacrifices to iteration times for safety, not even for performance. It's optimizing the wrong thing and making the wrong trade offs (for games), and probably the primary reason there are so few Rust games. Your character controller, dialog system, inventory management, renderer, physics engine, and 99% of your actual game systems don't need memory safety.

Multiplayer and file IO would benefit from memory safety, and it would probably be a good idea for existing C++ game engines to consider adopting Rust frameworks for those parts (even if it wouldn't be bullet proof).

With that said, game dev is an art/craft, and people should use whatever tools they like to create their art. It doesn't bother me. I'm just saying that, from a strictly engineering opinion, I don't think Rust is the most pragmatic choice for game dev.

[–] SorteKanin@feddit.dk 2 points 5 hours ago* (last edited 5 hours ago)

game systems don’t need memory safety

I see what you're saying but I don't know if I agree on this point. Games are buggy messes before undergoing a long and arduous quality assurance process. While Rust's iteration speed is maybe not great, I don't think it's really significantly worse than C++ iteration time, especially not with efforts like subsecond, that bevy uses for instance. And Rust is much more ergonomic and convenient to write than C++. It is in some ways a higher level language than C++ while managing to still allow for very fine optimisations, which are sometimes required in games. I think this would translate to productivity improvements too, although I don't really have any data to back that up obviously.

I personally think the momentum is mostly what C++ has that Rust is lacking - i.e. a large body of game developers who are already well-versed in C++ and a large and established body of game engines and frameworks for every kind of game. Games are often pressed for financing, so venturing into the territory of new engines and new languages is not something most game devs can afford or want to spend their money on.

[–] mitchty 1 points 5 hours ago

What iteration time sacrifice? I use bevy but I can just hot reload libraries if I need to. But honestly even debug rebuild times aren’t that bad. Methinks thou doth protest too much.

[–] bug@mastodon.gamedev.place 3 points 14 hours ago

@SorteKanin @entwine Rust is great for games even without engines, I think game developmers can learn to do stuff more low level. There is a great Rust wrap/binding to Godot. That and there are actually tons of little ways to make games in Rust. It depends on how granular a person wants to be! The improvements lately to the SDL3 bindings alone mean many things are possible. :) It's been way better than C/C++ so far for me.

[–] AngryClosetMonkey@feddit.org 7 points 1 day ago

The goal was to create a project where people can explore and learn Rust. When the time came to decide what we would actually do in this project, participants wanted to build a game.... So I set out to find a Rust game engine and determined that Fyrox was the best fit.