#highfidelity — Public Fediverse posts

There's a lot of talk about male-defaultism these days. You know what else is totally male-default? Avatars in 3-D virtual worlds. At least if they can be modified in-world instead of being monoliths like from places like Ready Player Me.

Almost everytime an avatar system is being designed for virtual worlds, it's designed with only adult males in mind. Eventually, someone will ask the question, "Well, and how do we make female avatars?" It'll probably be those who design either the avatar parts if the avatars are sufficiently modular or the complete avatars if they're monolithic. But this won't occur before the avatar system is finalised to the point at which it can't be substantially reworked anymore.

If you're lucky, female avatars get a somewhat slimmer waist. Or a pair of boobs, maybe only one that's being hinted at by a single bulge. If you're very lucky, they get both. If you're unlucky, there's only one hard-coded body shape, and all they get is more feminine clothing for the upper body, maybe even painted onto the one body shape that's available. In fact, you'll probably have as many "clothing items" for women as for men, all of which were designed by guys. All of whom are software developers with next to no sense of fashion.

Oh, and you get female-looking hairstyles, none of which are even shoulder-long. Well, except maybe for one ponytail that stands off so far that it has no chance of clipping into the body because the head motion is too limited.

Seriously, though: Even if your virtual world system is only planned for purely professional purposes, i.e. business, industrial, governmental, organisational otherwise, and extra care is taken that it will never be used by anyone in their spare time, even then this won't nearly be sufficient. If you plan to hold formal (not necessarily as in business formal) events, it'll be even less sufficient. If it's supposed to be an all-purpose virtual world system, a "metaverse for everybody" that people will use in their spare time, it'll suck completely.

I dare say that I've been using 3-D virtual worlds for longer than most of those who design them from the ground up nowadays. I've been in OpenSim for five and a half years now. I've built both male and female avatars. So I know first-hand what it's like.

OpenSim's default avatar, which also used to be the default avatar in Second Life long long ago, is female. She is named Ruth. But she's based on an avatar system that's mostly geared towards male avatars, and her hairstyle is more of a mullet because this avatar system can't even grow hair over the ears, because guys don't normally wear their hair over their ears. She isn't even really pretty. She can easily be switched to her male pendant, Roth: A bit less shapely, no boobs, therefore abs and pecs, the hair is shorter, and the face changes a little.

Now, if you want to see what kinds of avatars people actually make in Second Life nowadays, just look around (content warnings: eye contact, mild female nudity) PrimFeed, the Flickr alternative created exclusively for Second Life users. These pictures were actually rendered in-world and not by an AI. They show actual Second Life avatars, often even daily-driver avatars, in actual Second Life environments.

Short hair? Long pants? Almost only on male avatars, if at all. And today's male avatars have even more chiseled abs than 22 years ago. And chiseled faces and chiseled everything.

Female avatars, on the other hand, are shapely like you wouldn't believe from a virtual world unless you've been there yourself. Big butts. Big boobs and actual boobs. Dresses. Skirts. Bikinis. Lingerie. High heels, almost never under 15cm or 6 inches, sometimes even higher with platform soles. And: long hair. And with "long hair" I mean lush long locks, not just longer than male hair.

By the way: Nothing of what you see in the images was supplied by Linden Lab. Everything that the avatars and the landscapes consist of was made by users. It has basically always been the users who drove Linden Lab to refining Second Life's avatar system, often by working around its limitations and repurposing features.

Sooner or later, users with female avatars will demand three things for them. Male-centric avatar systems will be unfit to deliver either.

Long hair

The challenge with long hair is to not only make it look natural and, especially, keep it from clipping into the body. Extra challenge: If the clothes aren't simply painted on, keep it from clipping into the clothes when the head moves.

Of course, this point is moot if avatars can't move their heads. Which Second Life and OpenSim avatars can.

Skirts and dresses

This point is moot if avatars don't have legs. You know, like in Meta Horizon or (formerly Mozilla) Hubs.

But if they do, then even in a strictly business environment, I wouldn't too firmly count on all women being content with putting trousers on their avatars. They will wish for pencil skirts.

Skirts and dresses will pose a whole bunch of challenges. None of them is to keep people from upskirting, especially if the camera can move independently from the avatar which it should be able to do in good virtual worlds. In fact, depending on how a skirt or dress is shaped, preventing upskirting can be quite trivial. So that isn't one of the challenges.

No, the first challenge is to rig skirts in such a way that the legs don't clip through them, no matter how the legs move. If you manage to get that done with one kind of skirt, try again with another nine kinds of skirts. Including a pencil skirt.

Think you got that pat down? Well, here's the next challenge: Rig them in such a way that they look good when the avatar is sitting. This is rather trivial with pencil skirts. Now try it with a circle skirt in such a way that the skirt doesn't stand off as if it's made of wood. Extra challenge: Try it with a 1950s poodle skirt. These are just about as unthinkable in virtual worlds as 50s-style corrugated stainless steel diners. Trust me, someone will build the latter.

If you think it's smart to simply hide the legs from the skirt hem upwards so they won't clip, this will come back to bite you once the avatar sits down.

By the way: Even Second Life has never managed the former, and neither has OpenSim. And the latter is hit-and-miss with more miss than hit, and it works best with the old "painted-on" system skirt.

The cherry on top would be if skirts still flowed halfway naturally, and if skirts with a looser fit swayed with the motion of the avatar. This requires not only at least a basic physics model, but also a collision system.

High heels

Again, this point is moot if avatars don't have legs. But if avatars don't have legs, they'll be ridiculed. Yes, they will.

It doesn't matter how high the heels have to be. They don't have to be 15cm spike heels. Or 30cm spike heels with 15cm platforms.

So you want a virtual business environment first and foremost. Then your female users will wish for footwear that isn't men's leather shoes. And not ballet flats or Mary Janes either. Something with at least slightly raised heels. Just like they'll wish for pencil skirts; see above.

Okay, so you may manage to put high-heeled shoes on your avatar. Now, the first challenge will be to get the avatar's feet into the shoes. You know, the same feet that are firmly rigged into a flat position because that's all you need for male avatars (until someone comes and wants to cosplay Dr Frank N. Furter).

So you've reworked large parts of your avatar system to allow for other foot positions than flat? Good luck. Now adjust the avatar's Z position accordingly because the feet and the shoes are clipping into the floor. Ready to pump another few thousand man-hours into something else that you didn't take into consideration when designing your avatar system?

In Second Life and OpenSim, the former could only be solved in a satisfactory way by attaching different feet. Or, in the case of mesh bodies from Second Life, by giving it a whole bunch of feet in various positions and using a HUD to make the appropriate pair visible and the other ones invisible. The latter requires manual adjustment or some long forgotten trickery from almost two decades ago.

Finally...

I've read about four worlds and world systems that have tackled at least the former two points, namely by implementing a basic skeleton-based physics system that's a big upgrade in comparison to flexi prims in Second Life and OpenSim. Its big advantage is a basic collision model that makes hair and skirts (and anything else that needs it that you want to attach to your avatar) nicely flowy with little to no risk of clipping.

The thing is: Sansar was launched by Linden Lab as a kind of Second Life successor. That was at a point at which Second Life already had user-made, highly detailed fitted mesh bodies from various vendors.

High Fidelity was launched by Philip Rosedale. Also known as Philip Linden. The guy who invented Second Life in the first place and who led it for many years.

Well, and Vircadia is a High Fidelity fork, and Overte is a Vircadia fork.

This goes to show how virtual worlds are developed by people who have years upon years of experience with already existing virtual worlds and their users and creations.

And it goes to show that even reading Neal Stephenson's Snow Crash (which, by the way, introduced the word "metaverse" as early as 1991) doesn't give you the knowledge that personal experience with and in virtual worlds does. For Philip Rosedale has read it, and it has inspired him to make Second Life in the first place. But it has not inspired him to add certain elements for building female avatars right off the bat.

Unfortunately, however, even being prepared for that won't necessarily save a virtual world: Both Sansar and High Fidelity are no more. But that wasn't due to their avatar systems.

#Long #LongPost #CWLong #CWLongPost #SecondLife #OpenSim #OpenSimulator #Horizon #HorizonWorlds #MetaHorizon #Sansar #HighFidelity #Vircadia #Overte #Metaverse #VirtualWorlds #Avatar #Avatars #MaleCentric #MaleCentricism #MaleDefault #MaleDefaultism

Since a couple months ago, you can read it all over the place: "The #Metaverse is dead." Everyone agreed, because for 99% of all people out there, "Metaverse" refers to the series of 3-D #VirtualWorlds (to be) launched by #Meta, formerly #Facebook. And as far as I know, Zuckerberg actually tried to use "Metaverse" as the registered, trademarked, exclusive brand name for his worlds until he learned that he can't trademark something already used in a commercially published novel, namely Snow Crash by Neal Stephenson from 1991.

Thus, he settled for names like #HorizonWorlds which nobody knows nor cares about because everyone still speaks of Meta's worlds as "the Metaverse". And I guess people would continue to do so even if Snow Crash was turned into a massive Hollywood blockbuster with a budget of $400M that makes $4B in theatres within the first week.

What we can take away from this is that Mark Zuckerberg did, in fact, not invent the term "metaverse".

Oh, and just recently, Linden Labs started a massive PR campaign for #SL20B, the 20th birthday of #SecondLife which has also only recently started referring to itself as a "metaverse" to try and jump into the gap that the Horizons leave behind as Meta drops them like they're hot in favour of #ArtificialIntelligence.

Many have rubbed their eyes in disbelief. Didn't Second Life, like, shut down in, what, 2008 or 2009? Because the rampant news coverage about it died down back then. Yeah, but that was because it was no longer viable for commercial mainstream mass media to have virtual offices in Second Life after what few big corporations had joined it had left again. And when journalists stopped using their avatars (said avatars are still there, only unused), they didn't know what was happening in Second Life anymore. Besides, what was still happening in Second Life was only of interest for Second Life residents, but not for casual mass media consumers.

Nonetheless, Second Life continued to exist, and it does so until today. It even developed and advanced greatly. Today's avatars look nothing like those from 2007 when the hype was the biggest and from when the most images and videos seem to have survived. Oh, and they blow everything that Horizons has ever dared to demonstrate clean out of the water while consisting entirely of user-generated content.

What we can take away from this is that the Metaverse (no capital M here) is not dead, and that #HorizonWorldsIsNotTheMetaverse and has never been "The Metaverse".

However, between Snow Crash and the renaming of Facebook (the corporation) into Meta, the term "metaverse" was still used a lot, only it was used in places which next to nobody even knew, which are still largely unknown today. I'm talking about the worlds based on #OpenSimulator, a sort of free and open-source implementation of Second Life, and its community.

To give you a few examples: Alternate Metaverse counts as the fifth-biggest #OpenSim grid by active users and the sixth-largest by land area. It was launched in late 2019 under this name. That already was well before Zuck implied having invented 3-D virtual worlds. And the name wasn't chosen to cash in on Snow Crash, but because the word "metaverse" had been all around OpenSim for years already.

The Infinite Metaverse Alliance is from 2016, if not even older. And it has always been all about OpenSim with two grids of its own, one named Metaverse Depot.

#Metropolis, launched in 2008 was one of the first OpenSim grids, it was the first mostly German-speaking OpenSim grid, and when it was shut down for good almost a year ago, it was the third-oldest still existing grid. Its full name was "Metropolis Metaversum" for which there's proof from as early as 2010.

I'm tempted to say the earliest uses of the term "metaverse" in conjunction with OpenSim go back until even earlier in 2008 when OpenSim introduced the #Hypergrid which federated grids much like Fediverse instances are federated: For the first and so far only time in the history of virtual worlds, it became possible for avatars to travel between separate worlds with separate operators. Some said the Hypergrid was worth being referred to as a metaverse.

This was when it was increasingly attempted to define what a metaverse or the Metaverse is. Another idea was that "the Metaverse" refers to the entirety of all virtual worlds, regardless of whether they're connected or not. It would include 3-D worlds like Second Life, There or the various OpenSim grids, it would include 2½-D isometric worlds like Furcadia, it would include 2-D worlds and maybe even text-only worlds, and it would include out-right games like Minecraft or World of Warcraft, even if the worlds in the former are created procedurally. Basically, "metaverse" became the new "cyberspace".

And then there were those who had probably read Snow Crash and who knew what the Metaverse in that book is: a centralised, monolithic, corporate-owned walled garden. Essentially, that Metaverse was a vision of an Internet that had evolved into a 3-D world, but in 1991, the Internet largely consisted of corporate-owned walled gardens such as AOL and CompuServe itself, and Microsoft tried to establish its own one. That was three years before the World-Wide Web.

So while the requirement of being corporate-run and even a walled garden wasn't pursued further, "metaverse" was defined as being one single world. According to this definition, there isn't "the Metaverse", but there are many metaverses. Each OpenSim grid would be its own metaverse. No wonder not few grids actually refer to themselves as metaverses.

Sometimes, another criterium is added to the definition: It's only truly a metaverse when it's possible to move between separate locations (rooms, spaces, lands, call them whatever) by natural means. Usually, a virtual world has to be divided into smaller units, especially if these smaller units can be run by someone else than the creators/owners of the whole world. Now, this criterium means that these units have to at least be able to directly border on one another. An avatar standing near the border between two units must be able to look into the neighbouring unit. And in order to enter the neighbouring unit, the avatar must be able to walk or ride a vehicle that's actually moving instead of being a teleporter in disguise (I've seen both in OpenSim). Teleportation must not be a requirement out of basic technological limitations.

Now, imagine a virtual world that's IRC or Discord ported to 3-D just like the Metaverse in Snow Crash is AOL ported to 3-D, a world that only consists of separate, enclosed chatrooms which are built in-world as virtual conference rooms which you enter by logging into them and leave by logging out again. It probably doesn't have any windows. It definitely doesn't have a door working as such; either there is no door, or the door is decoration, or the door is the logout button, but there's nothing outside that door. If your avatar runs into that door, provided your avatar can walk and isn't bound to a chair at the conference table (yes, there are virtual worlds in which avatars can't walk around), it'll log out of that conference room and back into a kind of lobby. By the above criterium, this cannot be a metaverse.

However, if the door actually opens, and your avatar can look and walk through it into a hallway, from there into the lobby and even leave the building, then we're getting closer to a metaverse, probably even more so if the conference room is actually a separate virtual location operated by someone else than the lobby and the hallways.

Second Life fulfills this definition. You can walk around the mainland for hours, constantly crossing from one sim into another, all rented and designed by different residents, even though they all run on the same server cluster under Linden Labs' control. Sure, you can teleport, but that's only necessary if there's no other way to get somewhere. That might be because your current location and/or your destination is too remote, i.e. isolated by empty regions with no sims running in them which can't be crossed, or out of convenience because your destination is too far away.

OpenSim grids fulfill it, too, while the Hypergrid doesn't. The Hypergrid requires teleportation because it connects separate worlds and not different places within the same world. Otherwise, it's like Second Life while sometimes taking the "separate places with separate owners" part even further: Between renting land on grids and running a whole grid of your own, you can host your own sims and have them attached to certain existing grids. As a visitor, it might actually happen that you walk not only from one sim to another, but onto someone else's machine.

Still, if you look around, if you look at the various platforms that have "metaverse" painted on them, whether they're operational or only vague concepts, each one of their creators has a different definition of what a metaverse or the Metaverse is, always corresponding on what they plan their worlds to be like. Corporations that place all their bets on #VirtualReality claim that "pancake" worlds which can be accessed through conventional devices with 2-D screens like Second Life or the OpenSim grids can't be metaverses. Those who want to include the real life and #AugmentedReality or #MixedReality claim that this is part of the very definition of "metaverse" so that they can also deny VR-only platforms such as #VRchat or #RecRoom any metaverse status. At the same time, even companies that offer nothing more than e.g. concerts in virtual reality claim that their secluded concert venues make up a metaverse, too.

Corporate definitions of "metaverse" almost always amount to, "A metaverse is what we call a metaverse; all metaverse definitions by our competitors are false, they don't have/work on true metaverses." Exceptions are limited to Meta ("We're inventing the Metaverse from scratch. Wait, what do you mean, we can't trademark that word?") and Linden Labs ("We've had a metaverse before any of you even had computers. And our very own Philip Rosedale has actually read Snow Crash. Your arguments are invalid.").

Sometimes the definition of "metaverse" even goes hand-in-hand with a declaration of what makes a virtual world, and what's necessary to build and operate one. Cryptobros, for example, insist that the Metaverse/metaverses/virtual worlds can impossibly function without a blockchain, a cryptocurrency and NFTs. Others who invest in AI currently state that virtual worlds won't and can't be possible without AI. Second Life has been proving them all wrong by successfully and continually running a virtual world without a blockchain, without crypto, without NFTs and without AI for two decades now, but they build their business model on their customers either never having even heard of Second Life or believing it was shut down before summer 2009.

The IEEE even has a scientific paper on the definition of "metaverse". No, really.

This leads us to a set of criteria for the Metaverse or a metaverse that may or may not be valid.

The first one is that it's 3-D. This is easy to agree upon unless pre-3-D worlds protest against that definition.

Persistence is another criterium. The world must not only exist on your end-user device and start up when you join it and shut down again when you leave. This is generally fulfilled. Generally because many OpenSim users run their own grids based on the #DreamGrid distribution on Windows computers at home. Some do leave them running 24/7, others only start them up when they're at home and awake. And then there are those who only own one functional computer which therefore serves as both the machine they run their viewer on and their grid server. Now, the typical Windows user starts up their machine when they need it and shuts it down when they're done. So there are actually public grids that are only online when their grid owners are, even if that's only two or three hours a day. But this only applies to a limited number of grids and not OpenSim as a whole. That said, even grid servers in data centres running larger public grids have to be restarted every once in a while.

Thirdly, some make a functioning economy an absolute requirement for a virtual world to call itself a metaverse. Second Life has one that works so well that Linden Labs makes more money per user and month than Meta, all without privacy breaches. It helps that nearly all in-world content is made by users, and Linden Labs doesn't take offering free content in larger quantities kindly.

Its younger open-source sibling, OpenSim, however, which has been referred to as a metaverse or multiple metaverses would fail this definition. It's technically impossible to implement an in-world economy both with "monopoly money" and with virtual currencies that can be exchanged with real money, either grid-independently (Gloebit, Podex) or grid-specific (like #Kitely or #WolfTerritoriesGrid handle it). But the vast majority of grids has chosen not to include any method of payment for anything. OpenSim in general doesn't even need an economy because most grids by far are run by hobbyists in their spare time. And openly for-profit grids are not only suspicious, but usually not very long-lived. In the meantime, OSgrid, the first, oldest and largest of all grids, celebrates its 16th birthday next month (I guess), and it's non-commercial and running on donations.

By the way, OpenSim also took over Second Life's set of item permissions. But since so many avatars in OpenSim have access to admin mode ("god mode") which can override them, they're symbolic at best and useless at worst.

Immersion is a point that's being debated. However, this lastly depends not only on the underlying technology, but also on how in-world places are designed. Immersion is something that I personally am very very interested in. But most OpenSim users neither know what it is, nor do they care, especially not if it stands in the way of convenience. For example, building an in-door club with no doors to the outside saves the sim owner the effort of a) cutting a hole into the walls of the building and b) scripting and configuring a door. Sim owners tend to believe that if they wouldn't use such a door, nobody would. But a building with no doors is not very credible and realistic, and having to teleport to get into it and back out is not very immersive.

If we're talking about "the Metaverse" instead of single virtual worlds as metaverses, decentralisation is of course important. Now, by this definition, everything else from Second Life to #Roblox to #Fortnite to Horizon Worlds is just a bunch of centralised walled gardens and not even close to being part of the Metaverse. The few exceptions are all not corporate-owned; they include the #HighFidelity fork #Vircadia, the Vircadia fork #Overte and OpenSim's Hypergrid. The latter is made up from hundreds, if not thousands of separate grids, and very very rarely do even two have the same owner. On top of that, there isn't even an "official grid" run by the developers; lead dev Ubit Umarov only owns one standard region that's externally attached to OSgrid.

On the other hand, OpenSim entirely runs on one and the same software product. Even if various versions and even a number of forks are in use, it's only one platform and not several. And besides, how can the Hypergrid be "the Metaverse" if only a tiny minority of the grids that make it up pass the "metaverse litmus test" themselves because they don't have an economy?

Not even Vircadia could comply with this definition. It's decentralised, and it's commercial. Also, it's said to be fully compatible with Overte, so we already have two different virtual world platforms interacting. But for one, Overte is still a Vircadia fork, a soft fork even, so they aren't as different as Second Life and #ThirdRoom, and Overte messes with the economy requirement by being decidedly non-commercial at platform level already.

But seriously, debating such details is kind of futile as long as it's even unclear if it's "a metaverse/multiple metaverses" or "the Metaverse". So no, nobody has the privilege of having that one single "official" definition of "metaverse".