Compare and contrast: <a href="https://aframe.io/" rel="nofollow">https://aframe.io/</a><p>A-frame has got a certain amount of traction, it's component based and declarative. It feels like an extensible "VR HTML" (or "VR Web components"). React is similarly component based - but one step removed from the markup.
I don't understand the intent here. What am I building that's a browser VR experience? That's as foreign to me as saying a browser cooking experience. Those are words, but I just don't understand how they'd operate together in a way that would be enjoyable and easy to use.
The future of the VR web: everyone who uses it gives patent indemnity across their whole portfolio to Facebook. Facebook only gives reciprocation across the narrow patents covered by React/ReactVR, not their full portfolio.<p>If you build a part of your business on this, they can pull the license if you sue them for infringement in some completely unrelated area and leave you screwed.
I really want WebVR to work but when I tested with a Chromium WebVR-capable build I ran into a bunch of aframe samples that claimed my browser was not VR capable, some that were really choppy in performance, and a handful thay worked correctly.<p>After that I started thinking I should just build in Unity because there was a much greater liklihood of people actually using it with a client they download that actually works rather than hoping they would have a working browser for WebVR.<p>I also would like to have browser windows available inside of VR for interfacing with remote desktops or any existing 2d interface. Which you can do that with Unity using a browser component from the Asset Store.<p>But I would prefer using JavaScript/A-Frame if it would actually work for most people.<p>It seems that Altspace maybe is some custom build of Chromium build because they support aframe.
So, where is all the real work being done?<p>Oh, that's right, by THREE.js inside the normal WebGL API.<p>This is just a fancy way to execute a banal-ly simple 3D scene that takes a single parameter - what text to show. It doesn't do much and I don't see any useful abstractions for the usual 3D primitives and doesn't meet the requirements of 3D programming in general.