Hacker News Reader

> Decentraland, Voxels, and Horizon look like video games from 1990. There are people running around saying this can't be done. They're wrong. I've written a metaverse client that yields GTA V graphics quality... Users need a midrange gamer PC (NVidia 1060 or better, at least 6 CPUs).

Well yeah. Horizon Worlds runs directly on the Quest 2, not on a gaming PC.

Obviously anyone can achieve GTA V quality if you're powering it with a gaming PC, just by basing it on Unreal Engine.

It's a bit disingenuous to act like you've achieved something Meta couldn't. Now if you can get GTA V quality to run natively on Quest 2 hardware, then let's talk.

> Metaverses need to look as good and play as well as AAA title games.

Where's your evidence for that? People happily play mobile games all day long that are nowhere near AAA, and they use social media too. There's zero indication at all that graphics quality is the problem. People adore Minecraft and that's not exactly known for its AAA graphics.

I agree. This is what the VR purists are missing: normal people are not going to spend about $2000 or more for a desktop gaming PC, and another $1000 for the VR headset and base stations. The setup even for techies isn’t easy and straightforward either. Wires are also no fun for VR

Regardless of how you feel about meta, it’s really impressive that they were able to ship a decent 6DOF, standalone, VR headset at $299 which is relatively easy to use for even normal people.

Going on a tangent, even after all this time people still don’t understand what the metaverse is. The metaverse is simply an AR / VR platform that has online access. That’s it. So this why Fortnite, voxels, and decentraland are NOT part of the metaverse, but Roblox, Minecraft, Rec Room, and Horizon are part because they all have VR support.

> Going on a tangent, even after all this time people still don’t understand what the metaverse is. The metaverse is simply an AR / VR platform that has online access. That’s it.

And that’s because Meta did a shitty job to tell people what it is! Have you heard or seen one of their ads? In Germany, they go like „the Metaverse will help doctors to do complicated surgery. It might be virtual, but the Metaverse will have a real impact“

It’s like pure essence of bullshit, on the level of those self-reordering fridges promised in the 90ies. If Meta would just release a single, convincing use case now, maybe things would look different. But they don’t, because they can’t, because there isn’t one. It’s a solution looking for a problem.

> base stations > Wires are also no fun for VR

I'd like to highlight these parts again. Even as someone who has a desktop gaming pc and all: Base stations suck. Wired VR sucks.

Having wires get in your way while you are in VR is such a shit thing. I would never get back to it, not even for a generational leap in graphics quality.

And base stations automatically mean you need far more space (or "creative solutions"), something people living in e.g. flats don't have. Or don't want to have. Sure, I could probably rearrange half a room around it, but VR is not that great and probably never will be. Almost nothing would be.

> VR headset at $299

Wasn't Facebook losing money on each headset sold, though?

The same is done for consoles with the expectation being they'll recoup their losses through software sales. The problem is that you need to offer a compelling experience first, which Meta has failed to deliver on.

It’s $399 for the same headset now with the price increase, and it’s still the better overall deal.

> Obviously anyone can achieve GTA V quality if you're powering it with a gaming PC, just by basing it on Unreal Engine.

Unreal and Unity do much preprocessing and optimization of game content. All that takes place long before the game runs. Much of the smarts in Unreal Engine resides in Unreal Engine Editor, not the runtime. A metaverse client which displays dynamic user-created content from many creators doesn't have the benefit of that pre-optimization. There's not much instancing. Some content is badly optimized, has too much detail, oversized textures, or errors. You can't pre-bake lighting or shadows.

A metaverse client thus has most of the problems of an MMO client and some of the problems of a web client (browser). Games are normally pre-tuned so that the client does not face these problems. For a fully dynamic metaverse, content overload is the normal case, and frame rate has to be maintained independent of content changes. Think of what happens when you doom-scroll a web page. It's like that, but in 3D.

(There's a class of limited metaverse which works like a game level loader. Users build a game level, the level is loaded during a long preload period, and then you can use it. Sinespace, Breakroom, High Fidelity, VRchat, and to some extent Decentraland are in that category. You build your parcel offline, then upload, but you can change some object transforms in world and move things around. This allows some offline optimization. Clients for these systems are close to standard MMO clients. Dynamic stuff is tied to the frame cycle.)

Unreal could probably make their system into a metaverse-in-a-box product, with some functions that now run in the editor running on servers or as background tasks in the client. So far, they haven't done that, although Sweeney has talked about it.

High Fidelity isn't loading pre-baked levels. At least the original High Fidelity isn't, the new one is still a 2D Zoom of sorts as far as I know.

I'm part of the organization that picked up the old High Fidelity codebase.

Oh, good. High Fidelity was promising. Is that going anywhere?

Sorry but you're not even answering the quoted part of the person you're replying to in your reply. Their point is that games on the Quest 2 standalone don't look like AAA PC titles because of the hardware limitations, not because of the missing technical chops of the programmers.

That may be the technical reason, but end users don’t care about the why, when they can get superior visuals on even last gen game consoles.

Sure, that’s not VR, but I would argue that VR actually needs better visuals to be compelling to avoid an uncanny valley-like effect, unless you go super stylised and cartoony, which may work fine for many entertainment virtual worlds/games but good luck with real world brands and business.

Personally, I don’t even want VR, for two reasons: 1. it shuts me out of the real world too much, I don’t want to be that disconnected for long periods of time because I have coworkers, family, deliveries, my cat, whatever other things around me that I need to be aware of, and 2. in its current form factor, it’s simply tiring. I was playing with a PSVR2 recently and it was cool, but after about 45 minutes my face just got tired of wearing a headset and I wanted to do something else, even though I would have happily kept playing the game itself.

I should have said: I don’t want VR for metaverse/virtual worlds, outside of pure entertainment. I don’t mind the negatives I mentioned for short bursts of entertainment, but for “serious” use, work/business and multiple hours per day — no thanks.

>about 5% - 15% of the population becomes nauseated.

That's just those who can use VR at all. Three members of my immediate family can't because of eyesight issues, including one who's blind in one eye. According to the Quest manual, I have a condition that means I should contact a doctor before using it.

If I worked for a company pushing this, I'd have a hard time believing any claims they made about supporting diversity in the workplace. Masses of people won't be able to take part in a society that depends on VR for communication and collaboration.

>family can't because of eyesight issues, including one who's blind in one eye

Eyesight issues don't stop you from using VR. In VR you will have the same eyesight issues as you do in real life. If you use glasses, then you can still wear glasses when you use VR.

You have problems with your eyesight?

No problem! Just strap this blurry contraption over your existing issue to make it BETTER™.

That's exactly what prescription glasses are. I don't understand the point you are trying to make.

Are you trying to suggest that prescription glasses don't actually improve people's eyesight?

They don't emit light straight into the eye globe a few centimenters away from it.

The light passes through lenses made out of plastic or glass that are a few cm away.

Not the same, as any ophthalmologist will gladly educate you on the matter.

What is your point. Your eyes have no way of detecting the origin of light. When using VR there is no way for your eyes to know that the screen is that close.

My point is that you don't have any clue how my eyes feel with two tiny TVs a couple of centimeters away from them vomiting direct light with high intensity into them, and wide open pupils.

Not a single clue, and if you think natural light is the same, I suggest a new posture to watch TV.

Maybe you will learn something, maybe not.

>two tiny TVs a couple of centimeters away

Distance doesn't matter for light. It just adds a delay that is the distance divided by the speed of light

>vomiting direct light with high intensity into them

Most VR headsets are only 100 nits. That's less intense than than the intensity of light by being in an indoor room. Going outside is many magnitudes more intense.

>natural light is the same

What is "natural light"?

Your eyes physically bend to focus at different distances.

VR can simulate a specific distance, and your eye will focus on that, but the common current tech can't simulate different distances.

Your eye focus will be telling you it's a 12ft away screen or whatever, even when your binocular vision is telling you a flying tomato is right next to you.

Using tiny screens a few cms away from the eye makes a big difference, I am also not able to use it.

The screens are usually moved to be from 1 to ~infinity meters away by the lenses. Unless you mean that glasses are hard to use with some headsets because there is not enough room to fit. Lens inserts are typically the solution for that.

Definitly you aren't an ophthalmologist.

There are more eyesight than those which can be addressed with simple corrective lenses.

Then how do these people correct for it outside of VR? Your vision in VR will be similar to that outside of VR.

I'm not saying that all eyesight issues stop you from using VR, I'm saying theirs do. This is not theoretical, they have tested it.

Did you miss that I said one of them is blind in one eye? What glasses fix this?

>What glasses fix this?

One with a strong prescription. If you are fully blind in one eye you can still use VR. You will just see monoscopic just as you do outside of VR.

No it doesn't work like that. I'm amblyopic. I only have like 10 or 20% visual capacity on my right eye. Using those devices suck even more than a 3D movie.

Even if you close an eye entirely VR still works. I don't understand what your point is. If your eye only has 10 to 20% clarity than VR headsets are already clear enough that your right eye can't see a difference in clarity between real life and VR.

Except the real world is experienced with more senses than the eyes. VR is currently still visual-only. In the real world, any visual defects are worked around by the brain using other sensory inputs. No such possibilities with visual-only VR.

Other sensory inputs such as hearing and touch which current headsets and controllers output using speakers and haptics.

> including one who's blind in one eye.

Interestingly, even people who are blind in one eye can enjoy VR, including the perception of depth.

Binocular vision is only one of the ways we get 3D cues, and not even the main one. The main cue is motion parallax: how things move when we move our head (or body). This one still works even with a single eye.

I have read reports of people with one functioning eye enjoying VR. Just as in real life their eyes do not give a 3D image. Just as in real life, their brain can still infer 3D just fine.

>Fourth, there's really no role for "brands" in the metaverse. Despite all the hype, no real world brand has had any significant success in the metaverse.

IMHO, this is a crucial and often overlooked point. In order to gain popularity, a metaverse platform needs the sort of offbeat creator trendsetters who tend to push boundaries. Without them, the platform feels sterile and business-y. With the sort of content they will produce, however, family-friendly advertisers will run for the hills.

As Neal Stephenson wrote in Snow Crash (1992):

>You can look like a gorilla or a dragon or a giant talking penis in the Metaverse. Spend five minutes walking down the Street and you will see all of these.

Have you seen how much it costs to get an endorsement from Giant Talking Penis? Having that level of social clout means most firms can only afford to advertise in meatspace, because of the huge good will and consumer loyalty she commands.

> With the sort of content they will produce, however, family-friendly advertisers will run for the hills.

I mean, honestly, this in itself is one of the big problems with how our current society and economy work. The idea that most of our world should be "family-friendly" in the sense of censoring many perfectly healthy parts of human interaction and experience is toxic.

The current implementation, that often lumps most queer topics in with those that are "not family-friendly", is orders of magnitude worse.

> I've written a metaverse client that yields GTA V graphics quality and frame rate for a big virtual world. Talks to Open Simulator and Second Life servers.

That's cool, but...why would I want to use it? I think this is the fundamental problem that bullshitters like Zuckerberg can't actually solve. (And no, forcing employees to use it doesn't count.)

Good question.

When gamers try Second Life, they complain that it's s-l-o-w. That's a fixable problem for people with gamer PCs. They have enough CPU and GPU to do this right. The classic clients are single-thread and OpenGL, so they can't use modern gamer hardware effectively.

If you get rid of the sluggishness, user acceptance improves considerably. Or so says the literature on game design. You get flow. Performance and graphics quality alone are not enough for a win, but not having them means a lose.

I don't have all the answers. But I've pushed things forward a bit.

Lots of people are now making noise about metaverses and AI. Unfortunately, most of the people making noise are the same clowns that pushed NFTs. But there's some good work going on. This has the potential to make metaverses less empty, and human users will have NPCs with which to interact. It can definitely solve the emptyness problem.

NPCs are getting smarter. One of my NPCs in Second Life has a T-shirt that says on the back "Trainee - Someday we'll be in charge." That's less of a joke and more of a threat now.

At least for me the graphics and speed aren't the problem; the problem is that there is nothing for me to do in metaspace except maybe fly around and see other bored people flying around.

GTA V appears to be a very popular multiplayer game, so if a metaverse could have GTA V graphics and multiplayer performance then surely some GTA like activities could be implemented.

Which raises question of why people should not just simply play GTA, doesn't it?

I can play GTA on my Xbox for 3 hours. On an Oculus, my head hurts after 20 minutes.

But there appear to be no real use cases other than gaming?

And that’s a narrow subset of a subset of just single market.

It seems that Facebook was aiming at something bigger than that but they actually couldn’t even offer a coherent argument on what VR could offer outsiders of gaming.

> ... surely some GTA like activities could be implemented.

Running over people in cars? Who would be the players to volunteering to being ran over?

I mean, I don't play GTA, so there's that? And if I did play GTA I would... I don't know... play GTA, rather than playing a less interesting version of GTA.

Watch what people do in game lobbies like Fortnite, which are a bit of an open world: mindlessly destroy things, troll other people, show off superficial things like dance moves. A metaverse needs a bit more purpose, and constraints - a theme, or something that people strive for, or competition.

> Metaverses need to look as good and play as well as AAA title games.

I'm sure this isn't true. We've built communities using email, IRC, MUDs and web forums with more lasting power than any AAA game.

The main problem for Meta et al is that they want to sell ads (or something) to companies, and to do that they created a virtual world. Why would anyone want to take part in a world like that? If you want to build a world, base it on the notion that most people want to connect with others who are like themselves and people they look up to, and they want to "own" their place in the world.

We don't want to be passive spectators in a dead, plastic world created by megacorps.

I certainly don't disagree with your NFT rant :-) but do disagree with the characterization of the nominal "metaverse."

Back in the late 90's early noughts it Dave Rosenthal did a research paper on what it would take to create a "stadium" with 50,000 people in it where a sporting event was held. And create the "environment" that one gets when you are actually at such a stadium.

One of the challenges is that physics does a lot of work for you, so people across the field from you talking aren't just as loud as the people behind you talking. What's more, everyone in such an event would want "good seats" but at a real event there are people all over the stadium.

You could select people in groups of 50 and put them into the same "box" of good seats, and if you built a spanning tree overlay different people "near" each other, but it is kind of like the 3 color map problem in that you're mapping 50,000 people (probably in two groups of 25,000 for "home" and "away" teams) where they can "hear" the people around them cheering/talking/Etc. Dave felt faces were impossible although these days you could train an AI model using a webcam and taking pictures of your upper body and face from different angles and have the model then generate your faces of the people "around" you.

Annunciation and site lines are pretty straight forward renders, but the actual event (assuming it was really being played by real people) would require a combination of rendering and motion capture from the real event.

The amount of network traffic going back and for in n^3 to the number of people "near" you as they form a neighbor volume.

And that is just an event where you are a passive spectator.

World of Warcraft instance servers would do dungeon raids of 40 players (in vanilla I don't know if they do that any more) and their challenge was data base updates for damage dealt, damage done, and environmental changes. Of course most of the play field is fixed for that very reason but a more realistic one that was deformable would require more physics again. We got that with the Phys-X engine and others in the early 2000's but they have yet to be able to spew what they are modelling to the other players so that everyone has the same view of what is going on around them.

All this misses the most important point; Nobody will pay $500 for a standalone high quality VR headset to experience a shitty VR "Stadium" ticket that should cost $30 if you go to a real damn stadium.

Nobody fantasizes about living a virtual life that emulates real life. They want to live a fantasy life. Plenty of people managed that with just mid 2000s era WoW. Metaverse has nothing to offer

"Nobody" is carrying a lot of weight there :-). Do you have any reasoning here for your point of view? (not criticizing it, just wondering how you got there given what we know.)

For example, a ticket for the 49ers on the 50 yard line is ~ $1200 - $1500 depending on how important it is as a game. Also, existing renderers for sports figures are quite good given the market for video games featuring them, so it isn't clear that the visual quality would not be "just as good." Finally, the ability to do SLAM for stadium events is so good they can paint graphics on a moving player in real time as it happens.

So this is how my reasoning about it goes; Let's say it costs you $1000 for super hi-def VR goggles.[1] And the NFL decides to charge you $200 to sit on the 50 yard line "virtually" [2]. Your initial investment is $1200 ($200 less than the actual seat). Presumably you and your buddies would be allowed to be "next to each other." So there you sit, as a group, in great seats watching the game how you want to watch it, and all your friends are chatting with you. Every time you do this you save another $1200 off the cost of doing it "for reals."

But how can we even know that something like this would even be fun? The answer lays in the enjoyability of watching a movie "together" in VR. I think this is something of a sleeper as it isn't "a thing" in many people's mind yet, but it is definitely better than "simulwatching" while you have discord running. I suspect "twitch watching parties" would also have a similar vibe.

Other than the sex in Second Life, just hanging out is a very popular reason for participating in a MMO type universe. Having your "idealized" avatar with your face morphed on to it with all of its expressions retained intact and I think that would be pretty powerful.

So that is why I think it could be a thing, but I'm interested in your take on it as well.

[1] https://store.bigscreenvr.com/products/bigscreen-beyond

[2] https://theathletic.com/4522408/2023/05/16/nfl-average-ticke...

“Metaverses need to look as good and play as well as AAA title games.”

Strange that this line has been misunderstood when ppl write that (of all thing) email has done fine in community building without looking great. I think it’s the perspective - metaverses expect us to be _inside_ looking around, not looking at it from the outside. Who hasn’t thought, while playing an AAA title game, “aw man, this would look so cool if I was IN THE GAME!” So with that perspective it makes sense that finally being offered the chance to “be inside the game” is disappointing when it looks terrible.

> Who hasn’t thought, while playing an AAA title game, “aw man, this would look so cool if I was IN THE GAME!”

I haven't, because every great game I've ever played has made me feel like I was in the game. Even games with primitive graphics or text-based games. The actual VR games I've played don't achieve that.

Email has done well because it was part and parcel of general internet adoption. Email is also free, whereas a standalone VR set is not.

With respect to the metaverse being boring... I agree. Most VR experiences are lacking. This is a bit of a chicken and egg situation (in my opinion only: I know how bearish HN is on anything VR). There aren't enough users to get enough investment in creating experiences, and not enough experiences to really get the users... There's the odd exception, like Alyx, but creating a whole metaverse is a much bigger challenge.

User creation is one option but you identify the limitations of this.

One thing that might change the situation is AI creation. Considering generative models can already create 3d models, creating VR worlds should be soon possible. Being able to produce large amounts of VR content quite easily could make something like a metaverse suddenly feasible.

What was Facebook doing with all that money if it wasn’t trying to build a universe?

I once saw an Nvidia project where AI created 'medieval' buildings, but I can't find the link. I did find these:

https://developer.nvidia.com/omniverse/simready-assets

https://www.nvidia.com/en-us/omniverse

*edit: Rapidly Generate 3D Assets for Virtual Worlds with Generative AI https://developer.nvidia.com/blog/rapidly-generate-3d-assets...

Add another real problem:VR is just not that much better. My biggest takeaway from owning an Oculus Rift wasn't how incredible VR was, but how transporting not VR is. For experiences VR is super cool, but for playing games VR is worse in almost every way. Instead of instantly starting up and pausing if you get interrupted, you have to get the headset off and on. You can't pay attention, even slightly, to anything else while using a headset. That means no keeping an eye on children or animals. Using VR with a strong eye prescription is still a dilemma between a difficult adjustment or ordering very expensive custom lenses that also make it harder for anybody else to use the headset. Instead of just using an analog stick to move around and adjust where you're looking, you have to physically turn your head.

Frankly my TV is just a better gaming experience.

I might have agreed with you except... Half-life Alyx. I could be immersed in that one for hours at a time. No nausea or pain or anything and very well worth finding solutions for someone to watch the pets and kid for a few hours so you can play.

I haven't played a ton of titles, only own a Quest 1, but I found Superhot and Beat Saber to be distinctly unique and more enjoyable than non-VR counterparts in some ways.

On the other hand, I tried out some VR titles that just weren't very fun.

I think this is an issue of game design. Once the "wow" factor gets old, games still need to hold their own

What I love best about VR is that instead of puppeteering an avatar with a joystick, you get to do a lot of these visceral actions yourself. How this translates in real is that I no longer need to do a traditional cardio workout. I can now get the same work done while playing VR video games. I’ve lost about half 15 lbs so far

Of course for anyone who is not fond of moving or exercise due to health or other personal reasons, they’re not going to like the need to move for VR for now. I’ve seen some potential near term solutions though

As for the face bucket of isolation problem that you’ve cited, the latest gen VR devices have nearly solved this problem. Most AR focused devices like the nreal or meta quest pro are more open instead of enclosed. I have peripheral vision because it doesn’t hug my face. I’m guessing that Apple will also solve this problem within 1-2 years. Video games was just an afterthought for Apple Reality which is a good sign

It doesn't change your conclusion, but VRChat's Steam numbers are only about 40-50% of the entire population, however 50-60K peak per day is of course not massive, unless you are talking VR.

Even during the bussiest day of the year, new years, it is only close to around 100K total.

> Third, there's the VR nausea problem

VR nausea wouldn't be an issue if they focused on making the thing accessible without a VR set to begin with. With the exception of a tiny minority of early adopters, people are not going to buy a VR headset nor even bother with the setup if they have no killer app to use it with.

Immersion is pretty much the point of VR, and until Elon's Neuralink gets their shit together (not holding my breath) VR headsets are the least bad option we have.

Only a very small part of game or world immersion is technical though. The best VR setup in the world won't help you if the game is terrible.

> VR headsets are the least bad option we have.

True, but they're still a bad option.

Pardon my ignorance .. but what exactly .. or technically .. is the metaverse? how does one go about building a client? Are there api docs to refer to? is there like an equivalent to the 3gpp documents defining the standard of 5G?

"The Metaverse" is a marketing term looking for something cool to be attached to.

> I've written a metaverse client that yields GTA V graphics quality and frame rate for a big virtual world. Talks to Open Simulator and Second Life servers.

That's interesting, is that public?

There are test users. I'm not promoting it here.

Drat. Well, I'd love to see it if there's some way to do that.

Is there Linux support?

> I'd love to see it if there's some way to do that.

Email info@animats.com

> Is there Linux support?

Yes.

I don't know what's next either, but it seems like the things you describe are not terribly existential?

Technical problems and VR nausea will both be solved inevitably as hardware gets better and cheaper — which it always does, thanks to the magic of capitalism. There's also a lot of that magic happening in the world of GPUs right now.

The boringness seems like a game design problem. Someone will crack it sooner or later. I agree new mechanics are needed, and it'll surely take some time for something to emerge that takes off. This will also get easier as hardware gets better.

The "brands" thing seems the least concerning of all. Where the people go, the advertisers will follow. I think, in some way, this will also be solved by people figuring out good game mechanics for the Metaverse, too. Once there are incentive structures in place, advertising will find a way to slide in.

> There's also a lot of that magic happening in the world of GPUs right now.

I wish. It's not happening from NVidia.[1] NVidia will sell you an awesome GPU for an awesome price, but price/performance is advancing very slowly. NVidia's current entry level GPU card for desktops is $400. Also, NVidia's laptop GPUs that dissipate 130 watts are kind of a problem.

> The boringness seems like a game design problem.

Yes. We might see solutions from VR natives. See Phia's Virtual Reality Show.[1] Phia, in her early 20s and into VR for years, talks with insight about both gear and the social aspects of virtual worlds.[2] So far, though, no major breakthroughs.

The Yuga Labs / Otherside people tried making their virtual world a very expensive and exclusive experience to get into, in hopes of getting a non-loser crowd. But that turned out to be a Make Money Fast scheme that didn't deliver. Now the SEC is after them.[3]

[1] https://www.techspot.com/review/2206-geforce-rtx-3070-laptop...

[2] https://www.youtube.com/watch?v=xg_XB8It240

[3] https://www.bloomberg.com/news/articles/2022-10-11/bored-ape...

    which it always does, thanks to the magic of capitalism

LOL sure, the Invisible Hand will reach right into people's ear canals and rearchitect their vestibular systems using the power of THE ALMIGHTY DOLLAR

I just don't have any interest in a fake world. (Especially one suffuse with advertising.)

> which it always does, thanks to the magic of capitalism.

Just a general note: this isn't always the pattern the magic of capitalism gives us. There are as many examples of the reverse of that as supporting it.

As a quick example: toasters. The magic of capitalism has made sure that actually good toasters no longer exist in the consumer market.

> Once there are incentive structures in place, advertising will find a way to slide in.

Which, in the end, is what will keep me out.