I don’t feel that his opinion is all-encompassing. If Rust is being used in the name of security and reliability, then C/C++ should remain king of game development, where those two aren’t as important.
Only if you're using Unity / Monogame / other custom game engines that use C# as their scripting language.
Unity (IL2CPP) and Monogame (BRUTE) transpile the C# (byte)code to C++ to be able to deploy to consoles and the C# version they support lags several versions behind.
I don't think we can call C# "the new King of gamedev" until it is officially supported by console makers but it seems unrealistic for a couple of years.
Hopefully C# 7's NativeAOT fixes this (Sony & Nintendo will probably need to ship a runtime for C# for this to happen), I don't want to deal with transpiling stuff, I just want C# support out of the box.
As far as I can tell this only improves things on the Unity side.
What I have in mind is just making a game that can run on PS5/Xbox/Switch without using anything other than C# itself (and the platform API). No engines.
I vaguely remember proof-of-concepts/experiments on Nintendo Switch that made it possible to deploy NET Core AOT builds, but it was in some gitter chatroom so sadly I can't find it again.
Anyway, we still have at least a decade before C# or something else with GC takes over. C# is good enough, but something similar to Swift would make me happier.
If you're not using an engine, with a good editor, then you're already firmly among a niche few. The asset pipeline has been the dominating concern in game development for decades now, folks who willingly toss all that aside to write a game from scratch are among the fringe.
> If you're not using an engine, with a good editor, then you're already firmly among a niche few.
Not really. SDL, SFML, Monogame and FNA (and more) do not have editors, but a lot of popular indie titles are shipped with them.
> folks who willingly toss all that aside to write a game from scratch are among the fringe
Check out Dear Imgui. Indie shops such as exok (makers of Celeste) and AAA companies such as Rockstar Games (GTA) use it to create editors for their games.
You don't really need an engine. You probably need a framework if you're going cross-platform. If you're exclusive, you don't even need that.
If we start depending on proprietary engines instead of platform APIs, we're shooting ourselves in the foot in the long-term.
I'm aware of options like SDL, Raylib et al; despite that, indie is still dominated by editor-heavy tools, like Unity. I've been shipping games for nearly 20y; indie and AAA, and good Editors always win.
Even programmer-focused pipelines tend to evolve into designer and artist focused pipelines over time. It's simple: Dear Imgui doesn't solve the problem of rapidly iterating on an asset from Blender/Maya/Houdini; or from rapidly iterating on level design and scripting; or rapidly iterating on sound import and design; or... Over the course of development many indie studios end up reinventing tools that are already cheap or freely available.
Game developers generally persist hand to mouth. Rust isn't popular with us simply because there isn't a Unity or Unreal that offers Rust as a first class offering, and so it's up to the rare few who are interested in building engines from bare to build a new one.
Companies are definitely juicer targets, but I'm cynical enough to think that it's largely because it's not the game developers money that is at risk of being stolen, the incentives don't line up to encourage good security practices.
What's the asset pipeline story? Is there an editor, even?
What it looks like is another Raylib, Allegro, Ogre, etc; an engine these days can be expected to have tooling for artists and designers first and foremost.