The fact that there still isn't a good VR analogue of the web is baffling to me. I don't mean a browser in 3d space, but a standardized, decentralized protocol and runtime for VR experiences. It's kind of unfathomable that it hasn't happened. [1/n]

Back in 2006 or 2007, I started work on my first attempt to solve it. It was XMPP-based, designed all around multiplayer interactions. Frankly, it was a terrible Second Life clone, but Decentralized . I've done a couple iterations since then, but never got anywhere. [2/n]

I love working on @GrinningSoulEmu and I think it's going to help a lot of people, but more than that, I hope it lets me bootstrap a company. One that lets me play around with emulation research (my hobby) and eventually build this VR analogue. [3/n]

The fundamental structure of my idea hasn't changed in the ~15 years and neither has the codename: Microcosm. Much as you use a browser and it talks HTTP to a web server and gets back code and resources and renders that, you use a browser/viewer to talk to a Cosm. [4/n]

Unlike a web browser, though, your browser would have its own Cosm server providing your Home. There, you could customize your avatars, decorate, host friends, install apps, etc. All in a standard form with strong security constraints. This is, in effect, building a new OS. [5/n]

But because it's just another Cosm, you could decouple the two. You could write your own viewer/browser but borrow someone else's Home server, or vice versa. Users could mix and match too, but I don't think the average user would. [6/n]

The thing is, designing this system isn't *that* hard. Neither is building any one component. It's the fact that you need a whole ecosystem for it to be useful; much like a new OS. Embedding a browser is a cheap hack that would help, but at that point there's not much new. [7/n]

This is what I see as necessary for a real MVP of this idea. Maybe some can be very primitive or scrapped entirely, but I'm pretty sure most of these are hard requirements to even get a 1% market share among internet users. [8/n]

Protocol + standard for code (CIL?), a browser, a Home, a single-user Cosm (no "multiplayer"), a multiuser Cosm, a 3d editor/texture editor, useful frameworks for building apps, a search engine, an ad market (sigh...), a game engine and editor, a payment system, [9/n]

A web browser, libraries for automatic level of detail adjustment (the same Cosm should work on phones up, typically), a marketplace for objects/apps/tools/avatars, avatar builders, and so much more. Jesus. This is the first time I've enumerated them. [10/n]

Now, not all of this has to be built by one team or even one company. Hell, it *shouldn't* be. But how do you even get enough people to invest their time and resources into building this? I feel like there are only two reasonable options: [11/n]

Option 1: Be CEO of Google/Facebook. Choose this as your direction, build the core and then spin up teams to build the most valuable pieces. This is exactly the approach I don't want to dominate this kind of platform, but well... It's an effective one. [12/n]

Option 2, the Linus approach: In your naivete, drop a semi-working codebase on a mailing list for fun, and get lucky when more people than just you care about it. This can't be built by one person, but it could be started by one. The ecosystem can evolve itself. [13/n]

I genuinely think that something *very* much like this is the future many of us have been waiting for. But it feels like no one is building that future. I know people are building VR worlds, but that just feels like building Second Life 2020. [14/n]

While everyone else focuses on building MMOs without the swords, I want to focus on building the new HTML. Tim Berners-Lee didn't build an app, he laid the first piece of the foundation on which our modern world now stands. Who's going to start that in 3d? Maybe it's me. [15/15]