WebAssembly, WebGPU, and WebXR are the new VRML and will pave the way for an open, platform-agnostic metaverse where sites becomes worlds, hyperlinks becomes portals, profile pictures become avatars, and profiles become personal homes or spaces others can visit.
Interestingly, Meta's CTO just let slip yesterday about a web version of Horizon:
In the future, you won't just have to watch an overly long video or scroll three quarters of the way down a story to find a recipe, you'll have to travel through portals to different worlds!
There seems like such an assumption that everybody is aching to decorate new spaces and show off for its own purpose. How often do you visit a space just to check out the space, versus to do something in it? And that's especially true when pushed now by companies thinking it's monetizable, where now you've lost the benefit of not being constrained by real estate and physical material prices. So you've got advantages over video calls for keeping up with people far away, but what else?
> In the future, you won't just have to watch an overly long video or scroll three quarters of the way down a story to find a recipe, you'll have to travel through portals to different worlds!
And then one day Google will create little info cards that are snippets of the recipe VR sites. You'll still have to swoop and fly your way there but it'll be a smaller space.
There was some pressure to get SecondLife open-sourced and federated, but IIRC it didn't really get anywhere.
On the other hand, projects like Croquet/Qwak/Cobalt have been open-source for over a decade; are federated/P2P; already have "hyperlink portals" like parent commented; support existing standards like XMPP, VNC, etc.; have been ported to Javascript; etc. And yet, they seem effectively dead :(
The SL client ("viewer") has been open source for a very long time, and there used to be a thriving ecosystem of third-party forks. Nowadays Firestorm [0] is the one everyone uses.
The server software remains closed source. OpenSim [1], a community-driven reimplementation with federation support, has been around for a while but as you say hasn't really gone anywhere since the original wave of metaverse hype died down.
I think second life would be much bigger now if they didn't milk the longtail for so long. They charged based on costs from 15 years prior, and didn't add ways to easily customize things because it would upset the people making money selling 3d assets.
Both of which are really symptoms of the same overall problem, being unfriendly to new users.
It's a true federated system; you can run your own server, you can have portals to other servers, and, for grids which sell things, there are competing payment rails.
Open Simulator has a lot of spaces, but not that many users. Like most federated systems, it's a niche. Works OK, hard to use, few are interested.
Second Life continues to plug along. Right now, 50,844 users are in world. Which is more than any non-game metaverse. (Not sure about Meta; they don't give out numbers, but 20,000 has been mentioned.)
MMO games are far bigger; check Steamcharts. The top games are in the millions.
Don't believe any number about a virtual world you can't check from the outside. Concurrent users right now is usually the only honest number. There are systems which claim huge numbers of users, but their definition of "user" is "they, or some bot, put an email into the signup form."
The "metaverse" hype has produced a bit of growth, as people find out that most of the hyped systems are either nonexistent or very low rez. Usage is very low. Decentraland is around 1000-2000 concurrent users. They got up to 2600 once. Cryptovoxels is smaller. So is Sandbox. Sominium Space is in single digits.
that list is horribly outdated - so many of the grids in the list dont exist
Most people these days run their own grid through the dreamgrid installer. that actually increases fragmentation
and then the testing grid , osgrid, is down for multiple days every few weeks, which alienates and frustrates new users . overall activity is down, despite the pandemic. and my experience is that it's very very hard to keep users interested because it is lacking the critical mass of people. It also suffers from very high levels of drama
i find it very hard to believe that decentraland will keep such high numbers for long, it has received enormous coverage in the media eventhough it's bordeline a scam (imho). opensimulator has not received any of this
I dont think metaverses should be compared with mmo games - they are different things and attract nonoverlapping crowds.
Browser VR has been doable for years already. Three.js has been the de factor standard API. I have worked at a bunch of places that genuinely tried really hard with some actual practical, commercial uses for interactive 3D experiences and none of them stuck.
The Wikipedia example shows something kind of JSON-like but originally VRML was like HTML (SGML).
Something similar to that is A-Frame VR which is built on WebVR.
There used to be a way with at least one VR-enabled browser to navigate without exiting VR (if the other site supported VR). Not sure any still support it. I think this may be because it goes against walled-gardens interests. https://github.com/immersive-web/navigation#api-proposal
Interestingly, Meta's CTO just let slip yesterday about a web version of Horizon:
https://www.theverge.com/2022/4/14/23025899/meta-horizon-wor...