Yeah, there are a lot of VRML replacements coming out right now. X3D[0] is the official VRML 2.0, but I believe it requires a plugin (well, there is a THREE.js importer, but at that point you really should be using glTF[1] or THREE.js's own JSON format). There is GLAM by Tony Parisi[2], who actually did a lot of work on the original VRML. And very recently, Mozilla released A-Frame[3], which is trying to also build an ecosystem of reusable Web Components. All interesting in their own right. I think A-Frame has the best chance of succeeding, in that it has the most high-profile backing right now.
That said, they're all pretty low level. I specifically do not want users doing wholesale scene design in a hand-written script format. For the work I want to do, that's what a 3D modeling package like Blender is for, and as hard was Blender is to use, it's a lot easier to get a non-trivial scene working with it than hand-writing XML code. My Live-Editable 3D Environment demo is more of a stop-gap for a much more drag-and-drop experience I'm planning for the future.
The goal (and I'm certainly not there yet, but I've been making pretty fast progress) is to make something more akin to a Window Manager for VR. In the same way that DWM on Windows or X on Unix provide a higher level of abstraction for GUI elements beyond raw graphics primitives, I want to be able to provide a set of GUI elements that are useful for adapting HTML forms in VR. With Primrose, you shouldn't have to model a button that a user can click unless you have a specific need to completely redesign a button. There should be a button already that you can place on your scene/form with the same ease you would an HTML button in the DOM.
Long-story short, I'm trying to make the VB6 of WebVR.
It won't on a desktop, unless you're running one of the dev builds of Firefox or Chrome that have the WebVR extensions built in and enabled. It should work on mobile (though it will be better with Firefox Nightly or Chrome Dev). I don't know about iDevices, though.
Well, I managed to find an iPad 2 back in the deep recesses of one of my closets. I found a defect where my code failed to consider iPads as "mobile" devices, aka "has motion sensor".
Once I fixed that, there was another issue where iOS's version of WebKit doesn't handle orientation values in the same way as Chrome on every other platform (with Firefox having yet a third way, it's a very poorly specified API). Hopefully, Apple will eventually get on the WebVR bandwagon, because the WebVR API is much, much more consistent than the Device Motion API across platforms.
I don't have any iPhones around to test against those, but from the nature of the problem, I would guess you also had an iPad, because the detection should have worked for iPhone.
Perf on the iPad 2 was abysmal, which is not terribly surprising, considering the cross-section of Apple being hostile towards WebGL and me being not very experienced in optimizing draw calls.
I'm not trying to sell licenses directly on the site anymore (and had forgotten to change the README), but I'm still going to require copyright assignment, for many of the reasons stated in that SO thread.
[0] https://en.wikipedia.org/wiki/VRML