How does this compare to Mozilla hubs or the Third Room.
TR is/was:
* Open decentralised platform for spatial 3D collaboration, end-to-end encrypted by default thanks to being built on top of Matrix
* Web native built on WebGL/WebXR; no proprietary Unity or similar. glTF native for content.
* Apps written in WASM (compiled down from C, JS, and in theory Rust or Zig), using WebSG (our proposed web standard API for manipulating glTF content)
* Triple-buffered engine that can get 120fps on typical hw, including full immersive VR and XR on Quest Pro
* Direct manipulation in-world editing and inspector (“web style”), exposing the same model used by WASM apps.
* Networking is full mesh (so limited to 20-30 players per room), but these days could scale much higher by using the Element Call SFU (or by tiling the world)
Unfortunately, despite the Creator Update kicking ass in June (https://matrix.org/blog/2023/06/07/introducing-third-room-tp...) we failed to lock down long-term funding for Third Room and the team disbanded in July. The code and servers prevail though, and hopefully some day it may live again.
The original Croquet was a major inspiration for TR, and it’s cool to see the name live on. Hope they have more success than we did with TR.
a) metaverse hype collapse causing blowback oversteer against any kind of virtual world platform
b) the team wasn't interested in running it standalone as a for-profit startup outside of the Matrix Foundation, and Matrix doesn't have enough $ to fund core Matrix, let alone Third Room
c) big organisations who claimed they were interested in funding it within Matrix entirely failed to follow through (which probably maps to metaverse hype collapse). My favourite example is Rolls Royce, who ran a fullblown innovation challenge in May, saying they were specifically focusing on interoperable 3D collaboration: https://iuk.ktn-uk.org/opportunities/rolls-royce-sme-innovat.... We invested significant effort in applying, and never even heard back from them at all. As far as I know, they silently ditched the innovation challenge without telling anyone who'd applied to it. You can see our showreel for it at https://youtu.be/cFKJ-IuGaKA.
In terms of business models: selling SaaS hosting and proprietary apps & DLC on top of the open platform is the way to go, imo. https://thirdroom.io/preview#what-can-you-use-it-for has ideas for problem domains.
If anyone reading this happens to have $500K-$1M/y available to fund core dev, there's a small but finite chance of finding a way to get the project back on the road. But the core Third Room team has disbanded, and as per https://matrix.org/blog/2023/12/25/the-matrix-holiday-update... the focus of the remaining Matrix core team is currently monomaniacally on core Matrix foundations rather than future-facing stuff like Third Room.
Needless to say, I feel incredibly frustrated that we funded TR so far, and then had to stop before it got to escape velocity: simultaneously helping the open metaverse while setting it back too.
If one were to create a MVP for a collaboration platform. Would it be possible to do that with the current state of Third Room in your opinion?
Something like forced quantization into point clouds, 3D tiles, and gaussian splats would be a game changer.
They are original monochrome 1978 Space Invaders, and this is Fortnite with 10,000 players online at once, in full immersive VR?
OK, look, yesterday, I spent an hour carefully writing a simple explanation for some zeeb on Fesse Bouc on why you couldn't expand a ZX Spectrum with more RAM.
He replied with a one-line dismissal saying it was boring.
I blocked the nasty little troll.
Do you really want me to try to explain this, or is this another idle snarky comment?
Trying to summarize what's important about this tech is going to take a substantial chunk of my day. It's all in the article anyway. Have you RTFA and read the Wikipedia page on the original Croquet? Have you, in other words, done the absolutely minimal effort?
But when everyone uses pseudonyms, it is not obvious.
"Arathorn" is Aragorn's father in the Lord of the Rings and it doesn't tell me that anyone is a programmer of anything at all. How was I to know?
Yes, and thanks for doing so at https://news.ycombinator.com/item?id=38803004. I was hoping for a comparison with Third Room particularly, given TR was directly inspired by the original Croquet (albeit using WASM for shared computation rather than SmallTalk semantics). But it looks as you may have dismissed TR out of hand; I tried to explain its novelties at https://news.ycombinator.com/item?id=38773575.
> Have you RTFA and read the Wikipedia page on the original Croquet?
Yes, yes, and I used the original Croquet extensively in the mid-2000's.
> Have you, in other words, done the absolutely minimal effort?
Yes, I think so. I was more wondering how you felt modern-Croquet compared with other Croquet-inspired decentralised collaboration/comptuation environments like Third Room; I think I have the answer loud and clear now.
I apologise.
I was very impressed by the tech demo of Open Croquet that I received, but there is only so much space in a Reg article and I didn't feel that I was technically qualified to go deep into how the tech works.
I can respect good tech or a good demo even though I have no use for it myself. I don't do online gaming at all and have not done so since its very early days. It is outside my ballpark: I personally am not interested in using 3D, or gaming, or VR, or programming any of these things.
Which is also why I do not follow other entrants in this field. So I am sorry but I'd never heard of TR, or of Mozilla's offering.
I'm interested in development for a complete web based experience using WebGL 2.0.
Never heard of Third Room before and I can't find anything much about it, so I can't help you there.
That’s the reason Brian Upton (Rainbow 6 etc) joined Croquet originally, and used it for his “wide wide world” experiments.
Anyways, you can also do things naively. Some RTS games just handed down the entire state to every client and let them figure it out. You can see how people would take advantage of that to strip away fog of war during online matches.
I know nothing about Mozilla Hubs or the Third Room. No comment or comparison to make.
This "Croquet" is a new version of a circa 20 year old project: https://en.wikipedia.org/wiki/Croquet_Project
That Croquet was an effort to extend a then circa 25 year old project, the Smalltalk programming language,.
Smalltalk is more or less the original object oriented programming language (there was Simula before it but it has different goals). Smalltalk aimed to make it easy and fun to create GUIs and GUI-driven, 2D graphical, software.
Smalltalk largely flopped in the market although some commercial and FOSS versions survive. For instance the Eclipse IDE is an offshoot of one of these.
Some of the original Smalltalk team created Squeak, a modern colourful Smalltalk that is entirely FOSS. Squeak has been used to create lots of other things such as the children's educational programming tool Scratch.
Croquet was built on and in Squeak. Croquet extended Squeak into two new directions at once: 3D graphics, including hardware acceleration, and transparent networking.
Croquet could be used to build 3D apps and entire 3D environments, including 3D games and VR along with lots of other things.
But also, at the same time, the idea is that Croquet instances on different machines can talk over a network, using a protocol called Teatime, so that those instances are connected and can share data, live, in real time.
The idea is, as I understand it, that it makes stuff like multiuser video conferencing so trivial it's basically a no-code exercise.
It is trivial to build something like Zoom or Teams or Google Meet: you draw a room, place some picture frames on the wall, connect each one to the mic and camera of the local machine, and boom, you have a video conference in 3D, without coding.
It is meant to make much more complicated ambitious stuff easy. It extends Smalltalk and makes it able to do 3D as easily as Smalltalk did 2D stuff at Xerox PARC in the 1970s, and it makes it all network-transparent, so connecting instances is as easy as coordinating two Smalltalk windows to share data: literally a fun game for schoolchildren.
Xerox PARC's Smalltalk is the 1970s tech that was the original inspiration for the Apple Lisa in 1982, and a couple of years later, the cheap cut-down version, the original Apple Macintosh in 1984. That inspired Microsoft Windows, and Windows created an industry of cheap 32-bit PCs with integrated graphics and sound which is where Linux developed and thrived.
Croquet, in 2004, took this 1970s tech that predated (and caused!) the microcomputer boom and put it on the Internet with hardware rendered 3D.
Smalltalk is the original language that popularised OOPS and so which led to C++ and Java and Python and C# and .NET and Scala and Rust and Go and all that. It is also the language which inspired the Mac and GUIs... all GUIs.
Smalltalk also directly inspired later OOPS languages, such as Self. Self subsumes Smalltalk objects into "prototypes", using a Smalltalk-like syntax.
Self later inspired Javascript, which replaces the Smalltalk-like syntax with C-type curly-brackets syntax.
Now, OpenCroquet moves Croquet off Squeak Smalltalk and into Javascript instead. It is Croquet, ported from Squeak to Javascript, so it runs inside the browser.