I've used both and Horizon Worlds' lack of development is mind-boggling. It's positioned as the crown jewel of the Metaverse, the core from which Facebook will build the rest of the concept, and yet it's so barebones it's like no one is even working on it.
VRChat has completely user-made avatars. To support this, they have a robust content moderation system with a fair amount of user control, and bounded performance requirements. Horizon has very limited Mii-like avatars, but with even less customization. This means they don't have to moderate avatars or worry about performance, and also crushes the boundless possibilities of self-expression into a sea of identical corporate art style drones.
VRChat also has user-made worlds. So does Horizons. The difference is that Horizons worlds can only be made out of primitives (cube, sphere, triangle, etc.) This makes it easy to enforce performance requirements and also makes the entire metaverse look like a poorly developed PS2 game. It also runs worse than a well-optimized environment mesh, but apparently Facebook doesn't trust their users to figure that out. Second Life actually shipped with this system originally (in 2003) and later abandoned the system because of these and other problems, but learning from the past is apparently not in vogue at Facebook.
VRChat has a fairly granular safety system, with configurable boundaries, default permissions, friend settings, etc. Horizons has a half-hearted attempt at this but leans mostly on the ominous promise that Facebook is recording everything you say and do, and if someone reports you an unreachable Facebook admin will review your past actions for content violations. This works about as well as you'd expect.
These and other problems are, IMO, all bad decisions, but they're also low effort decisions and that I do not understand. I actually like VR and want Facebook to succeed here (hopefully incentivizing competition), and the department formerly known as Oculus is doing a great job with the requisite hardware. So why, why, given that the entire company has been bet on this, is the flagship Metaverse software of the world's largest social media company lagging behind a random startup making their metaverse in Unity with practically no money?
> world's largest social media company lagging behind a random startup
This seems to be a really common recurring theme in the tech industry and particularly the games industry, and I wish I understood why
Just look at Minecraft, Valheim, any of the recent Pokemon games... A single person or a skeleton crew can somehow always seem to outproduce gigantic companies full of engineers and artists on the same level as the ones from the skeleton crew.
I think this is because once you are a running on infinite cash, it becomes a no-brainer to apply what you've already learned to achieve that success. Also, stakeholders will require that. So pull in a bunch of management, MBA-s, designers, analysts, etc. What you end up with is lot of friction and no real innovation.
I imagine they're going with Mii-like avatars because they have business contexts in mind. What they could do instead is to give people at least two avatars for different contexts and then let the tool/environment set which they're using.
e.g., if you're in a business meeting, it automatically runs with your Mii-like avatar, otherwise if you're roaming a desertscape to relax on the couch after dinner, you can be a pink lemur in a cape and cowboy hat. Might as well prep for both contexts now rather than putting off either potential customer.
VRChat has completely user-made avatars. To support this, they have a robust content moderation system with a fair amount of user control, and bounded performance requirements. Horizon has very limited Mii-like avatars, but with even less customization. This means they don't have to moderate avatars or worry about performance, and also crushes the boundless possibilities of self-expression into a sea of identical corporate art style drones.
VRChat also has user-made worlds. So does Horizons. The difference is that Horizons worlds can only be made out of primitives (cube, sphere, triangle, etc.) This makes it easy to enforce performance requirements and also makes the entire metaverse look like a poorly developed PS2 game. It also runs worse than a well-optimized environment mesh, but apparently Facebook doesn't trust their users to figure that out. Second Life actually shipped with this system originally (in 2003) and later abandoned the system because of these and other problems, but learning from the past is apparently not in vogue at Facebook.
VRChat has a fairly granular safety system, with configurable boundaries, default permissions, friend settings, etc. Horizons has a half-hearted attempt at this but leans mostly on the ominous promise that Facebook is recording everything you say and do, and if someone reports you an unreachable Facebook admin will review your past actions for content violations. This works about as well as you'd expect.
These and other problems are, IMO, all bad decisions, but they're also low effort decisions and that I do not understand. I actually like VR and want Facebook to succeed here (hopefully incentivizing competition), and the department formerly known as Oculus is doing a great job with the requisite hardware. So why, why, given that the entire company has been bet on this, is the flagship Metaverse software of the world's largest social media company lagging behind a random startup making their metaverse in Unity with practically no money?