The original VRML is dead. Around 2001, a successor to VRML started called X3D (https://en.wikipedia.org/wiki/X3D), with VRML now being called "ClassicVRML". However, I haven't seen any active development on X3D since I started in VR seven years ago.
I know X3D was supposed to be more than just an XML format for a scene description. Like, I think there was supposed to be an object repository, and specialized viewer applications, all under the X3D name. The branding was classic early-aughts confusion of "name everything the same thing, cuz it's 'integrated'!"
I think X3D mostly suffered from being too much, too late. Most of the VR stuff you'll see in the browser now is either developed in Three.js or Babylon.js (or exported to one of those two). "Good enough, available now" won out. Plus, the whole anti-XML movement kinda put a damper on anything with "X" in the name (though X3D did start discussing a JSON serialization format, I don't think it ever left draft stage).
For the most part, people serialize scenes to GLTF files these days (or a Wavefront OBJ, if they're scrounging models off the Internet). GLTF has a compressible binary format that makes comparable scenes a lot smaller.
One might draw a rough analogy to GLTF => PNG, OBJ => GIF, X3D => Jpeg2000. It just never got enough adoption.
Oh man I remember VRML. Do you also recall WildTangent? Was also another 3D engine in the browser. At that time I was convinced the entire web would become a mix of VRML + 3D navigation. Websites would become mini 3d worlds. To visit the Contact Us page, you would have to move your character/camera to a room etc.
I've also some nostalgia for the demos produced in this time. I cannot for the love of god find this one where you are in some film noir type of detectives office you could walk around in. The lighting effect, ambient traffic sound from the window, it was magical around the 2000s.
This echoes my understanding as well. Virtual Maker uses BabylonJS under the hood for rendering. That said, I've been interested for some time in finding/developing a standard format for describing 3D content which can include some form of input handling (GLTF is just rendering, AFAIK). Think <button>, but for 3D/VR. Maybe then we could port some simple experiences easily between engines.
I think part of the problem is that the full understanding of a scene with interactions requires a graph structure, but most of our tools are really only built for editing tree structures.
I see this time and again with just regular, ol' 2D DOM and JavaScript. You can define your layout in HTML and then play whack a mole with bugs if elements you got via document.querySelector move or get renamed, or you can define your layouts in JS and pay a relative huge cost in rendering and complexity.
Plus there is the added wrinkle in 3D of having to be very persnickety about memory and shaders and order of operations. How do you make a declarative, semantic scene description that can figure out that this geometry over here can be batched, this over here can be instanced, and never the twain shall meet?
Ah did you then modify the BabylonJS editor? I've been thinking of this approach but worried people have been submitting issues of glitches and oddities that made it seem unstable.
I've been trying to write a 3d editor from scratch with Three.js and Vue 3. It's quite daunting task and I am more appreciative of the work Babylon guys are doing just wish there were more funding so they can knock some of the older issues out.
I know X3D was supposed to be more than just an XML format for a scene description. Like, I think there was supposed to be an object repository, and specialized viewer applications, all under the X3D name. The branding was classic early-aughts confusion of "name everything the same thing, cuz it's 'integrated'!"
I think X3D mostly suffered from being too much, too late. Most of the VR stuff you'll see in the browser now is either developed in Three.js or Babylon.js (or exported to one of those two). "Good enough, available now" won out. Plus, the whole anti-XML movement kinda put a damper on anything with "X" in the name (though X3D did start discussing a JSON serialization format, I don't think it ever left draft stage).
For the most part, people serialize scenes to GLTF files these days (or a Wavefront OBJ, if they're scrounging models off the Internet). GLTF has a compressible binary format that makes comparable scenes a lot smaller.
One might draw a rough analogy to GLTF => PNG, OBJ => GIF, X3D => Jpeg2000. It just never got enough adoption.