Is Dual Universe Closest to Being a True "Metaverse"? Virtual World Expert Who Helped Define the Term Says So -- And Not Sansar, High Fidelity, Etc.

After saying I was impressed by High Fidelity’s ability to get 350+ avatars in the same local metaverse space, a guy who actually helped define the metaverse told me that he wasn’t very impressed. Will Burns, Vice-Chair for IEEE's Virtual Worlds Standard Group, showed me this video from 2016, and told me to watch. It’s a tech demo of Dual Universe, an upcoming sci-fi MMO that's only been slightly on my radar up until now, and you should watch too:

“As you can see,” says Will, “they not only have 1200+ concurrency and able to scale, they also have a persistent universe with planets. I have to reiterate that last point -- an actual virtual universe that is persistent.” At least that’s what the developer, Novaquark, claims:

Dual Universe is the first game where potentially millions of players can interact together simultaneously, inside the same giant universe. We call it the continuous single shard. No multi-server player separation, instances or loading times, just one enormous persistent universe that never stops.

“Not that I want to disparage Philip Rosedale and his team, or even Ebbe Altberg and Linden Lab, whether this is Second Life, High Fidelity, etc,” Will went on. “But the honest truth is, 300 or 500+ concurrency isn't very impressive to me when I know what is possible, and generally how it is made possible two years ago. It's not very impressive when a hybrid decentralization method would ultimately shatter those paltry numbers.”

He brought that point up to Philip recently, who just suggested, Will says, that “I apply to work for High Fidelity." 

However: “[H]ere's the problem: that platform and others, who all call themselves a Metaverse, are all off on the wrong foot, and locked into their particular paradigms. Unless we're taking things like Dual Universe into account, it's just an echo chamber among the pretenders about their accomplishments and milestones. Lots of nice virtual worlds, but that's about it. While everyone else is waving around the ‘Metaverse’ title and debating who is better, or who is going to be the Metaverse, it's the shallow end of the kiddie pool by comparison to Dual Universe or even No Man's Sky.”

Will first helped craft a definition in 2008, inspired by Neal Stephenson’s original description in Snow Crash, which IEEE adapted in 2013. “Now, if you type the word Metaverse into Google, it pops up with that definition from Wikipedia, which in turn cites the IEEE Virtual Worlds Standard group for it.”

That started an interesting conversation worth discussing here -- and by the social VR industry as a whole -- but first, a caveat from Mr. Burns:

“It comes across like I'm salty or disparaging,” he says, “when I'm really not intending it like that. But it has to be said that we are amid a VR hype where nobody wants to look at things with a critical eye. I see a large group of virtual worlds crowning themselves as a Metaverse because it's marketing hype. But ‘a’ Metaverse sounds as silly to me as saying ‘We are an Internet’. What we have are very impressive (or in some cases less impressive) virtual worlds:

“Second Life comes out as a MetaWorld at best, but even struggles there. Sansar took the Worlds Inc. model, just like Blue Mars, so they are a MetaWorld at best, but isolated non-connected spaces perceptually. High Fidelity follows that same route and comes across like Cybertown or Worlds Inc back in the 90s. It's like we've all taken a large step backwards and harbored collective amnesia about the past while declaring our current progress as top of the line when in reality we're retreading old ground. Rehashing the past approaches with maybe a technical twist.”

Then compare the capacities of these new worlds with what’s happening technically with Dual Universe:

“[T]o me, watching things from the sidelines, despite spending twenty or so years helping to better define the Metaverse with a great amount of awesome people... it's like watching used car salespeople trying to tell the public that their Honda Civic is a Bugatti Veyron or the Millennium Falcon. It does a disservice to call every virtual world a Metaverse because it deliberately waters down what it actually is and could/should be. It comes down to: I expect better, and so should everyone who even remotely gives a damn about the future of The Metaverse.”

I point out that Dual Universe seems more like a traditional if ambitious MMO, but not a Metaverse in the Stephenson sense, where users could, for instance, buy and sell content for real money.  

“But Novaquark realizes this and calls themselves openly ‘A Metaverse Company’,” Will says -- and indeed, they do. “Much as the semantics make me twitch, I think Novaquark is the only company I'd give a free pass to on that usage of Metaverse. Because you could always use that architecture not as a MMO space game but as a general purpose Metaverse architecture... with those other inclusions like a marketplace and currency system, etc. If they ever did, it would put literally everything to shame overnight.

“I'm not saying they are the Metaverse. But I will say they and projects similar have the best chance of actually evolving into it. Whereas High Fidelity, Sansar, etc, even Second Life or OpenSim...they likely aren't going to evolve into The Metaverse. The foundation was wrong and retrofitting them just isn't feasible without an architecture and presentation rewrite.

“But that's not a bad thing. Identifying what we got right and what we got wrong is paramount. Then we can compile all the things we got right into a single initiative and have an actual blueprint for the real deal.”

So how would he start?

“Let's say you made planets that would be traversed, and even interplanetary travel means. Let's say you had resources to mine and collect. That items had a resource requirement. Suddenly, you have an organic explosion of jobs and needs to be met. Communities have purpose as cooperative initiatives. Trade routes open to transport and sell stuff to other areas.

“Suddenly, because you planned ahead, and built architecture to anticipate organic growth, you see cities and entire governments. Tons of resource opportunities and jobs. A real economy of scarcity. People have a reason to live closer together, cooperate, and so on. You then can drive a vehicle on a cross country road trip. Fly a plane thousands of miles. Charter a ship and explore the ocean. All that opens up. You have purpose.

“It all arises not because you arbitrarily placed them there. But because you created a system that fosters it all naturally. That's what I see in things like Dual Universe or No Man's Sky. Not that they are a Metaverse, but that they begin with the right premise to build on that soundly to create a virtual universe that fosters all those things naturally and not arbitrarily. Those I see as most likely to evolve into the Metaverse. Solid foundation and bigger picture thinking.”

To Will Burns’ mind, it’s too late for the companies calling themselves Metaverse creators to do anything like that.

“[They] aren't going to admit openly that they screwed up. They have millions of dollars in investments on the line. They collectively have to stay the course and insist that they alone are the next big thing, even if there is clear evidence they aren't.

“I do not envy Ebbe in this respect having to push Sansar. Or Philip Rosedale having to stand by High Fidelity like it's the next big thing. In the latter case, the reputation sold the idea. People are buying into the cult of personality more than the soundness of the overall assertions and plans.”

He believes too many people (I guess including me) are going along with their momentum.

“There's no room in those situations for a dissenting voice. Somebody to say that maybe the foundation was screwed up or that it won't be what they imagined it would be or sold it as to the public and investors. They have to see those to the end, and stick to it. They can't do a 180 to make those adjustments. They are locked in.”

I tell him it’s likely that Dual Universe will just be a modest cult hit, like Eve Online before it.

“Dual Universe will probably be a niche MMO,” he says. “But we're not talking about the initial use of that architecture. We're differentiating the underlying architecture from the use case of it. If you look at the tech on its own, it can be used as a general purpose virtual universe. Like Second Life on universal steroids. OASIS.”

A lot that Will Burns is saying is debatable -- for one thing, I'm not sure High Fidelity's concurrency load test can be fairly compared with the Dual Universe test: The former involved 350+ human-controlled avatars in a small space (say 75 square yards), while the latter involved 1200 computer-controlled avatars in a large-ish space (say one or two square miles), with avatars that are farther away given less rendering attention.

But to his basic point: Yes, we in the Metaverse industry (such that it is) should probably consider Dual Universe a serious contender -- perhaps succeeding where most are too inherently limited to go.

What's your take, reader?

Above: Will Burns with Linden Lab CEO Ebbe Altberg

Comments

[Pulsar]

Among the virtual worlds I know, Second Life is still the closest to a single instanced and contiguous world. Other platforms are closer to a 3D chatroom instead, where at most you would teleport from a "room" to another one (whichever they are called "rooms", "experiences", "spaces", "worlds" or whatever). In SL, instead of teleporting, you can travel, you can explore the world by driving a car, flying a copter or on a train or sailing... for hours. In Second Life there are even organized grid flights, social cruises, drives, with dozen of people traveling together. Sometimes you can also meet someone and make a new friend while you are traveling by yourself. Thanks to the progressive loading and some client-side interpolation between regions, traveling in SL feels like a single experience, apart only the infamous region crossings disturbing it. However, sailing on the Blake Sea with an optimized boat, you would barely see it.
Unlike Sansar, HiFi reduces the loading waits somewhat by showing the world progressively too, before everything has been loaded; SL seems to do it better, in HiFi you still have small waits, but it's already better than its competitors.

But if you disable the interpolation in SL, so that it stops at region crossings, usually people call it an experience breaker, even if the stop lasts for just a couple of seconds. What many hoped for - in the aviation, sailing and other vehicle communities in SL - was a world without region crossings. No, not even just that couple of seconds, interpolated or not.

Now it's true that we are going backward: not only these new ones don't offer that - they make you wait from several seconds to several minutes, up to an half hour, to change place - they don't even offer this somewhat contiguous world and kind of continuous experience made with regions. You only change room, whatever large is it, to an entirely different place. Indeed back to the '90s, and you can still visit Active Worlds today and see that it's kind of the same old approach.

[Pulsar]

Dual Universe is impressive. Star Citizen can go smoothly from orbit to the surface as well (since 2014-2015, I believe), but I don't know if SC can handle that amount of characters/avatars per time (and how many people are still trusting that project).
The cluster of cooperative servers mentioned in the video may allow a better load balancing. SL was conceived as having a different server per each 256x256m "region". Now you may have more regions per server, CDN, and other stuff, but essentially you can end up with a server choking under the load of a shopping event, while lots of its fellow servers just stay idle. If regions become just coordinates (so scripts stay compatible), and a cluster of server runs a "continent", that won't happen and crossing a region would be like moving anywhere else.
I agree with many things Will Burns says. Even if Dual Universe would stay a niche MMO and won't become a Metaverse, creating a new virtual world based on a similar technology would scale up much better, rather than the room-like approach (or the "sim" or region one) and you would have a more world-like experience than a room to room thing.
IMHO it's true that they screwed up with the new virtual worlds (if the goal was a Metaverse) and it's true that rewriting the architecture from ground up now would be a pain or kind of late anyway. It's also true that the famous next thing still doesn't show up, while reality is a while that knocks on VR folks' door. Maybe they can hope in some niche instead; after all IMVU kept doing well alongside SL. Unless someone really creates an OASIS-like metaverse that blows everything else away...

If now someone lays the basis of an alive virtual world, and designed to work as Dual Universe does (I don't mean DU, just the tech under the hood), and allowing to travel etc. but keeping a desktop mode too... well, I don't know how successful it would be, but I know I'd go there, especially if it's also creative. I don't think instead of something with resources to mine, etc. That sounds like just another MMOG to me, thus reducing the scope considerably. Even if you have a world simulator in mind. The Sims Online was launched just few months before Second Life, and was shut down in 2008. SL is still here. Maybe if it's something like a Minecraft on steroids and you mine "prims"... LOL dunno. However, when users can create their own world, they can create games too. Outside the Mainland, in SL there are games like that, roleplay mini-continents, like Amazon River, where you have resources to harvest, to mine, hunting, fishing, villages, canoes, quests, fights, socializing, and so on, with flying disabled, experience keys prevents people to teleports "magically" everywhere, you have to eat and you could drown underwater. That could be fun for someone. But if the whole SL was like that? Uhm.

[dahlia]

I've always had a lot of respect for the developers of EVE Online. I think their current concurrency record is somewhere around 54,000 but it could be more by now.

[Clara Seller]

I really enjoyed this piece and I think Will Burns is seeing the big picture.

The Dual Universe concept is inspiring and totally kills that "experience in a cracker box" vision being thrown at us.

Thanks to Will for speaking a little truth. It's refreshing. The amount of B.S. being churned from these "next big thing" hype machines has become freakish. There's simply nowhere to go, once you've buried yourself in so many lies.

[Darianknight]

I'd like to point out, Wagner, that the dev video itself states that a little more than 1000 players/clients were spawned in a few hundred meters area. Not a few miles. Secondly, those were individual clients running to simulate the load - they still were individual clients loading and moving around. So in the case about purely human driven clients, I guess the question becomes whether being a meat driven WASD button presser or having the system run the clients itself is important to the outcome of the network stress test.

I know that folks will often have strong opinions on this subject, and that's ok. What is and isn't a "Metaverse", right? My line of thinking is simple - If we're looking at something at the scale and ability of OASIS in Ready Player One, which closely interprets the more modern definition of Metaverse (as I provided to Solipsis, ACM and IEEE), then I would simply point to systems like Dual Universe as a candidate to evolve to that if the involved companies that have those systems are inclined to utilize them as a general purpose Metaverse.

In order to get to that point, you have to think in terms of vastness and emergent behavior. Everything else then seems to run on lowered expectations if you're holding it up to OASIS level... So we get these companies that say "Oh, OASIS just isn't possible" and then they adjust their lowered expectations accordingly to create whatever virtual world they are making. But if you point out things like Dual Universe, those lowered expectations no longer factor in... you begin to see that it actually just might be possible beyond what we have been told.

So it just took a group of developers on Kickstarter to ignore the naysayers, the status quo, and do what was said to be not possible.

Whether you think your particular virtual world of choice is or isn't a Metaverse comes down to a simple question: What do you expect the Metaverse to be? Do you expect it to be an amazing and fantastic journey int o a virtual universe or do you think it's a limited scope virtual world by comparison?

My perception of "The Metaverse" looks like OASIS, and seeing things like Dual Universe emerge gives me hope that something like that is possible for real.

[Susan]

Well, this really changes everything. Thanks for sharing this one, Wagner.

[Ryan Schultz]

Interesting! Thanks for posting this, Wagner. It makes me want to go investigate further.

[Lannorra Sion]

I am not familiar with this new game. I watched trailer, but it did not give any real information. I have a deep background of MMORGs, been playing for long time. So I just want to say this... What makes Second Life, Open Sims and other virtual worlds so powerful and addictive is the truth to "Your only Limits are your Imagination!". In other words, user created content. What has always disappointed me with other MMORGs that I played is I could only make what the Devs ALLOWED me to make. blah. I had been playing Star Wars Galaxies in 2005 with my son and I was a tailor, and the Devs would never give us anything we would post as the top items tailors wanted. It infuriated me. When I heard about this new "game" Second Life, where you could "make anything you wanted", I took a look and they hooked me from the first moment I uploaded a texture to make something that did not exist before. THAT is the true power of virtual worlds! And I think if Metaverse is another existance for a human then its important that we can create what we want. I am not creating things some group of Devs decided was good. I am creating what I WANT TO CREATE! Open Sims keep me, exactly because I can create worlds from MY imagination. If this new game includes texture, object, etc uploads, then I will agree it might be a better platform or whatever. If they do not... then they are just another game.

[Wayfinder]

I appreciated the candor in this interview. While Dual Universe is as-yet-unproved Alpha and a lot of claims and promises yet to be fulfilled... what Will says is spot on. Second Life is "successful", but could have been far more successful if more coding expertise and less stab-your-customers-in-the-back greed was involved.

Yes, Second Life is outdated code that can't really be "fixed". More than a decade of failing group chat (simple text?) is proof of that. Years of major bugs on the JIRA that are still unfixed is proof of that. The inability to cross a sim line with a starship full of people is proof of that. As comparatively impressive as SL is, the viewer code leaves no doubt it is spaghetti-coded patchwork that has painted itself into a corner.

I told Linden Lab years ago that if they didn't start stepping up their game and focusing on customer needs rather than stockholder demands, sooner or later a big-dog competitor would come along and eat them alive. Will that competitor be Dual Universe? There's no telling at this time, but it does seem like this company has the right idea: no postage-stamp sims, no sim-crossing issues, no "we own everything you create" mentality (at least not that I've seen thus far. I still have to read the fine print.) Above all, this doesn't seem to be a land-based economy-- which is one of the major flaws of most SL / OpenSim virtual worlds.

It will be interesting to see what evolves over the next few months as the next generation of virtual worlds hits the market... and does its best job of spinning Linden Lab in circles.

Wayfinder Wishbringer
Founder, Elf Clan

[Hamlet Au]

"the dev video itself states that a little more than 1000 players/clients were spawned in a few hundred meters area. "

I couldn't find a section of the video where we actually *see* that, and what it looks like. I'd want to see that, and see how performance holds up, for a more one-to-one comparison of HiFi stress tests. And yeah, human-controlled avatars are going to add hundreds of thousands of more variables into the simulation to process, compared to an automated one.

[Clara Seller]

We've given years to Sansar and Hi Fidelity to give us very underwhelming results.

I'm very willing give Dual Universe tech time to develop. A cup of it's promise is worth ten truckloads of gas masks, solitary confinement, and "kinda like WordPress for VR".

People want to be creative, empowered, pioneers, explorers, and free. I have no faith that these very attainable goals of VR will ever be perused by big tech. There is money and success in giving people what they want, but they don't.

"Think Different" was the tech world of the eighties. "Think Sheep" is the co-opted tech world of today.

[Matti Deigan]

I can't really take Will seriously after the last line on comparing Dual Universe (not even out, more game like than a social platform) and No Many Sky (Game, Proven Technologically) to not being in the kiddie-pool when it comes to Metaverses: Perhaps by his definition; but even reading up on that I feel like both of the examples do not meet that definition as using them as examples opens up other examples that have already done it in the public and are available and not closed:

First of all regarding the examples:

Dual Universe: The 2016 Dev Diary demonstrates 50 seconds in the networking tech with local clients.... Nowhere does it demonstrate having positional and rotation Audio chat during this, nor does it demonstrate other important factors to social vr which all the others mentioned are. Removing the Audio and Avatar Animation streaming (instead of playing from animations) that High Fidelity has from the equation, it is no different to MMOs, aside from its Proc gen and minor user generated content by existing assets. But if we do this, Dual Universe stuff is not impressive nor is High Fidelitys record;

Modern day MMOS do similar things already, and having 1000s of users on a MMO server without Audio is the norm: Eve Online, WoW, and Guild Wars 2 are modern (even if a bit dated) examples of having a single area where you have tons of users over a large landmass (or space), even if instances do exist in them. So what is demonstrated is just norm.

No man sky... Which only supports 16 players max in an instance, where only 4 are visible as avatars, and the rest as orbs... Even on release it didn't have this.... I can't even… Why was this used as an example for this case of comparing concurrencies?

Ok, Sure if it wasn't the concurrencies, maybe Will is aiming for more of the metaverse concept where you have a universe: But it is No man sky is not fully persistent, and it is still Centralized into a single server that spins up instances for clients to cross connect with.

Minecraft communities (Yogcast) have stuffed 2600+ users into a single server. And that's already an example of being able to connect multiple worlds together on a single server with teleporting between the worlds and services using portals…. Can that be then considered a metaverse as well? If so, isn't that more impressive than Dual Universe or No Man Sky, as it is already doing what everyone is trying to do.

But even so: it wasn't "as playable" but real usage always shows how Stress test is not indicative of real use case.

These worlds are also bound by the rules of the game and you can define rules per world. All the above examples do not go any further than that. Because if they do: then we can go even further:

If No Man Sky and Dual Universe can be considered as examples of metaverse, then it would be saying the Space Engineers or Elite Dangerous are also varying degrees of metaverses, or even Star Wars Galaxies when it was still alive. Those also have User Generated Content (to some degree, ED also has, but it revolves around the RP) but in all of those you can bouncing between various worlds/planetoids 'seamlessly' and able to do things on them.

So honestly, I can't really take comparisons of Social VR platforms to Games, because doing so opens up comparisons to MMOs. It’s apples and oranges.

The thing that made High Fidelity's 356 users impressive was the low latency audio, for everyone involved. that is 356 users cross talking, and everything communicating together on a single server. You could hear it as a crowd. This can be expanded upon and it is something even Mumble with positional audio cannot approach without having many many latency issues; The audio in High Fidelity is positional and rotational, so it takes account also the direction your avatar is facing. This is fairly important when it comes to HMD users, but to also facilitate HMD users you need to have joint data streaming, so that is also something that comes with what the server has to handle.

Audio in games is a funny one, as many gamers simply pop into using Mumble (TeamSpeak or Ventrillo in the past) or Discord. Even some vets of SL groan at the concept of Voice chat because it introduces too many variables, and perhaps they don't want to use it, because there are just too many issues to deal as people do not have good quality audio.

So Developers tend to overlook audio communications. But Communications, both Text Chat and Voice chat are a very important tools on Social platforms, and Metaverses, if Metaverses are to be where where multiple users interact with each others over across many worlds, across many devices, it is very important to ability to communicate regardless of the means..

For Dual Universe and No Man Sky, Voice comms simply are no considered: which probably the heaviest part of all server communications, which was also the point of the Test for the High Fidelity Stress test. So to meet a goal of the OASIS or Stephenson Metaverse, or OZ from Summer Wars;

You need to able to communicate through Voice, Text or even Body Language, for those in HMDs.

There is a lot of tech there that makes the entire subject very interesting, and this can be scaled once you apply to octree algorithm (which was also demonstrated in the Dual Universe) , that already is used in the Entities in Hifi to distribute users among many servers.

With Hifi domains being 16 km cubed, and still being able to create links to other domains without any issues (you can create portals that you can just run into to go elsewhere), there is plenty of methods to go around and able to eventually create almost seamless transitions between domains. The tech is there, but its absolutely rough.

[Aliasi Stonebender]

If concurrency all you need to be a "metaverse", World of Warcraft had all of us beat ages ago.

I don't consider games to be such, even though they are certainly a species of virtual world. If you cannot do arbitrary things with arbitrary avatars, in a world that can be constructed on the fly (not premade assets), then it fails in any comparison to Second Life as a "metaverse".

[Gaga_Gracious]

As for the definition of what the Metaverse is I certainly don't think it is anything like a Minecraft where one has to search for resources to make things, and that is what Dual Universe is from what I read and that is basically an MMO and not a Metaverse in my view. The closest we have ever come to a true Metaverse is Opensim where no one need be dependant on a proprietary platform owned by a commercial enterprise. Opensim is open source and can be used by anyone without obeying the self-serving TOS of a company like Linden Lab. And yet commercial Opensim grids can and do exist along side private worlds and the neat thing is they can connect via Hypergrid or choose not to, which ever. Now that, to me, is the closest yet to a true Metaverse. The rest are all walled gardens of one kind or another so I don't think it is as important what you can actually do in a virtual world to meet any definition of what a Metaverse is but how it is able to connect with other worlds and without obligation to any central authority.

Opensim is decentralized by its very nature and grid managers answer only to public opinion which they can pay heed to or ignore at any possible cost to their reputation as a grid. And if you run a commercial grid based on Opensim and want to attract users, renters and merchants public opinion can make or break you. Moreover, Opensim has completely independent currency in the form of Gloebit that can be used on any grid as well as in-world currency if the grid owner chooses. And finally, there is Kitely market that can deliver purchased products to any grid that permits it. Some grids even have Bitcoin!

I tend to agree with Aliasi Stonebender above who said: "If you cannot do arbitrary things with arbitrary avatars, in a world that can be constructed on the fly (not premade assets), then it fails in any comparison to Second Life as a "metaverse"."

A true Metaverse, IMO, has to be like the Internet itself where everything is interconnected by design and only walled off or separated by individual user choice. And whatever one wants to do in their world the tools to do it should make most things possible.

[irihapeti]

all the players in the vid are the same mesh character. When they can do 1,200 uniquely different mesh characters moving and interacting together at 24 FPS on individual hardware clients without crashing a client then they will have something

why I say this is because I can get just over 2,000 same character simulations on my Win64 client computer connected locally to my OpenSim server running on a laptop

350+ individual and independent avatars vs 1,200 same character simulations ? I go with the 350+

not to say that the idea of what they are hoping to achieve is wrong. It isn't. Just that they still have a quite a long way to go to be even in the same street as High Fidelity, Sansar or even Second Life with its 100 avatar cap per server

[Philiprosedale]

I discussed architecture and next steps a bit in my blog about the test here https://blog.highfidelity.com/356-avatars-together-ea8546e86279

Agree with others here that world-sized contiguous spaces with many people able to build together in then is the correct end-goal and definition of 'Metaverse' - that is the recursive architecture we are building, captured in this blog https://blog.highfidelity.com/high-fidelity-system-architecture-f30a7ba89f80

That's what we are building - 5 years work so far, going as fast as we can!

Remember whether you are talking about Dual Universe or High Fidelity - tens of millions of SERVERS will be required to support the people in the space. So the big design here isn't just technical it is also economic in nature - there needs to be a motivation for everyone to work together and contribute servers. Roughly speaking, it will require about as many internet connected PC's as we currency have worldwide to simulate/store/stream such a space. That's why we are also working on the blockchain-based currency systems.

[ProphetZarquon]

I think that it's still too early to call any of these things a potential metaverse; I'll explain why in a moment, but first I would like to point out that "concurrency" is being thrown around in a rather loose fashion in this comment section so far.

WoW does not have thousands of users on a single instance & that does matter, a lot. Whether it's an instantaneous portal, or a loading screen, or actually having to use menus to select another instance; unless there is a space where all the players can crowd in & see each other, then we're really not looking at anything innovative. All other essential aspects of a metaverse have been done, in one title or another. There is one thing & one thing only significant about DU, & that's the potential to have truly seamless travel where everyone shares a single space.

That said, DU can't (in current form) meet my definition of a metaverse, in part because it's a single space. As a game, it's strength is that everyone must work from one set of resources; one procedurally generated map that can be seen simultaneously by all. That strength (true concurrency within a single universe) is also what makes it just a game, & not a truly user-created metaverse.

Which brings me to several core assertions about what I think it means to be a true "metaverse":

1) Multiversality: More than one set of physical rules. Content that crosses genres. Future tech; ancient Greece; Shogun; cartoon logic; fantasy lands with flying dragons, where tech items cease to function. A true metaverse needs to be able to support any model of reality, ideally with as little waiting time when moving between as possible. For all its other limitations, this is something Second Life does fairly well. Fairly well: Those universe boundaries should not necessarily inhibit bringing content one to another; The aforementioned inability to fly a starship full of passengers across sim boundaries is a major downfall for SL. If a user wants to take their 150ft dragon to a spaceport or shoot down fairies with a laser cannon, that should be possible unless the local rules of a given location prevent it; but SL still does the multiverse thing better than any other title I've seen.

2) User-created content: SL's bread & butter. DU is trying to incorporate this philosophy to the greatest extent they can by release, but I wouldn't expect 30ft trolls walking around with steam driven cat-launchers imported from an .FBX file. Moreover, the game is a single universe with static physics rules; Users will not find fantastical fantasy realms of magic within DU. That said, I think (DU's developers) Novaquark are ideally positioned to become the preeminent "metaverse company", by licensing their innovative (unprecedented, really) all-inclusive-multiplayer server technology to other titles which can then offer a more open multiverse, replete with wildly different rulesets but potentially still accessible using a single user-side client. DU's initial (single) persistent universe does promise the possibility of user-created constructs & organizations of greater scale & scope than even Eve Online could accomplish. (From what I've seen, no other title - either released or announced to be in development - offers the potential crowd-size of DU.) In contrast to this incredibly large but finite-in-possibilities single universe of DU, Second Life has great flexibility in user-created content & is mostly lacking in performance & scope. You simply can't put 30,000 or more players in a single SL instance & have them all see what each other are doing (much less digging holes in the map). For the time being, I think SL is still the king of user-created, with Minecraft variants perhaps a distant second.

3) Concurrency: This is where DU shines. Given that the title is decidedly SciFi in genre & has no current plans to implement different physics or tech trees in different places, it cannot take the title of "best multiverse" in it's current from; but the server tech behind DU is another matter. I fully believe that if Novaquark is able demonstrate that their all-inclusive-multiplayer technologycan work at acceptable client-side framerates without bogging down the servers, their methods will become the new gold standard in beyond-massively multiplayer tech. If it works, I have no doubt that single-instance experiences will become one of the three defining characteristics of future "metaverse" development.

I haven't seen much to encourage me about the future of High Fidelity yet, but I do think that some kind of distributed server technology will be essential in creating the first true metaverse.

This brings me back to the first two requirements: Multiversality & User-created content; because I believe that the true future of the metaverse (& indeed the future of telecommunication) relies upon decentralized networking. The server loads of projects like DU, High Fidelity, Star Citizen {cough-it'll-never-work-cough}, & most other modern titles are astronomical compared to the user-operated game servers of old. Whether it's by blurring the lines between instances, or a novel implementation of Merkle hash trees, or network-usage tokens for distributed computing, or some other tech I've never heard of, I firmly believe that user-operated systems with no central authority will be necessary for achieving a true freeform metaverse.

Right now, we can barely get a few thousand users on one map; but to reach a vision of something as pervasive & widely used as OASIS, I don't think any single organization can be in control. In many ways, when we talk about a metaverse, we're talking about reinventing the internet; & make no mistake: The internet needs reinventing.

Centralized systems are great for extracting profits, but they're increasingly unsustainable from the perspective of user rights, trust, & waste. Cryptocurrencies have stagnated in part because profit-driven third-parties have usurped the development process. (Bitcoin is almost entirely wasteful, but other tokens need not be.) Mesh networking has plateaued at a level where telecoms use it extensively for backhaul management but end-users hardly know it exists; not because of real scaling limitations (try another protocol) but because it isn't profitable enough. In terms of freedom of communication, open accessibility, user choice, workflow resiliency, freedom of innovation, & equal opportunity, the corporate-operated internet is a miserable failure most epic in scope & impact. As someone who experienced the "wild west" days of the internet first-hand, my opinion is that the state of modern computing is so deeply corrupted & perverse that I hardly recognize it. It is a restrictive, profit-biased, usurious hellscape of predatory negligence where greed is encouraged, & accountability is lost in an endless shell-game. What we desperately need is a new network, built & operated by the people, for the people.

A true metaverse will need to be founded upon this principle & evoke its essence from end to end.

Only by establishing & upholding a new paradigm of ethical social computing can we hope to see anything like the virtual freedoms espoused by so many futurists, authors, & engineers.

TL-DR: Freeform creativity & bringing all users together is not enough; A real metaverse will require new ways of networking, from distributed hybrid-mesh computing, all the way down to empowering end-users to contribute more than just money & decorative content.

[irihapeti]

pretty much agree with Philip Rosedale on the general broad technicals design philosophy

when we look at this from the pov of the client viewer as a simulator

the pertinent questions are: How many servers can the client simulator handle gracefully ? and what content/activity do each of the servers handle ?

chat server, voice server, world positional server, physics server, texture server, model server, animation server, friend and group list servers, ban/block server, land/parcel control server, outfits server, client input device server, upload and download servers, asset transfer (blockchain) server, currency accounting server, etc etc

then next questions include: How many of the assets (models, textures, sounds, etc) can be cached viewer side ? How much of the animation and physics can be simulated client viewer side while maintaining integrity across all players in the client side view ?

then the overarching questions. How many control servers are needed to oversee/ensure the stable realtime relationships between the lower-level servers and the clients ? How many clients can be assigned to a control server ? How many control servers can be grouped relationally?

pretty much every multi-verse developer has been through this questioning thought process. And I think pretty much also everyone who has been agrees that the key to high concurrency multi-player multiverses is to look at it from the pov of the client viewer as the simulator and not the servers per se. Individual client hardware determining much of these parameters, on a realtime connected client by client basis

[Ai Austin]

I wrote a blog post back in 2014 as High Fidelity and Sansar were just being announced to note my own wish list for a next gen virtual world and metaverse of such environments...

http://blog.inf.ed.ac.uk/atate/2014/08/04/wishlist-for-next-gen-virtual-world/

1. Avoid Isolated Islands – Unified Single Instance Virtual Space
2. Avatar Identity and Names
3. Improved Procedurally Generated Flora and Terrain Features
4. Migrate to the New World – Pack your Trunk
5. Travel to Other Virtual Spaces – Pack your Suitcase
6. Content, Content, Content, …
8. A Viewer for the Metaverse

[LoirL]

Matti Deigan wrote: "Minecraft communities (Yogcast) have stuffed 2600+ users into a single server. And that's already an example of being able to connect multiple worlds together on a single server with teleporting between the worlds and services using portals…. Can that be then considered a metaverse as well?"

I consider Minecraft to be the first breakthrough virtual world, created by its users, a prototype-metaverse perhaps. A new generation is accustomed to playing, creating, communicating, watching, listening, and reading via Minecraft. That children and teens are having formative encounters with games, online social experiences, and user-generated content across media platforms through Minecraft suggests that the future for virtual worlds is robust. As youth grow accustomed to participatory imaginary worlds in which they can communicate with friends, customize avatars, create, play, share information, have fan experiences, follow characters and stories, compete in games, and strengthen media and computer skills in order to show off to and compete with their peers, they bring the elements of virtual world participation into everyday life. Since Minecraft emerged organically from single-player games into user-generated multiplayer synthetic worlds and a community-created virtual world, the experience of Minecraft is an experience of agency and empowerment.

The load tests in High Fidelity are significant because High Fidelity is working on a distributed model in which users own their virtual spaces, yet they are still part of something bigger and can be connected. It's the infrastructure of a metaverse, the crucial components of architecture and currency. So load tests are about pushing what servers can do, to strengthen the building blocks of what can become the metaverse.