Interesting use of the example image, I wonder if Apple is thinking of iMessage on the web, after we've seen FaceTime for the browser. I wouldn't have thought it a possibility, but it's strange they used that example.
Apple proposes adding "model" element to HTML that displays 3D content
111–120 of 241 posts
Re: Apple proposes adding "model" element to HTML that displays 3D content
#112Earlier quoted context omitted.
Proposing adding something to the HTML spec is not “creating a new standard no one asked for” Don’t be deceptive. Inaccurate generalizations and hyperbole add nothing to the discussion.
Holding a knife and moving your hand back and forth is not killing someone, either, but my omission of a few important details is the deception. In this case, the something being added to the HTML spec is a new standard that they are creating. I'm not going to argue whether anyone asked for it, but "nobody" is obviously hyperbole; Apple is asking for it.
Well there ya go, I called you on trite generalizations and hyperbole. Adding no truth or insight to the discussion.
I for one think this is a future proofing move with little current utility. However I’m not one who thinks the AR/VR/ 3D Models in browser bandwagon is going anywhere within the next 5 years.
Re: Apple proposes adding "model" element to HTML that displays 3D content
#113This doesn't really seem like the right abstraction. 3D models are inherently interactive.
There could be some value in supporting 3D video, where interaction is inherently limited to 3 degrees of freedom. There may even be value to the "videos" being represented as 3D scenes with animation i.e. rendered on the device.
However, "models" are something you need to interact with. To achieve an even remotely acceptable level of UX those interactions absolutely need to be custom tailored for the content.
Let's say you want to display a 3D building interior. If you're going to allow any form of free camera movement, you need to define how that logic works.
Do we want first person walk-through? If so, you need collision detection. For simplicity, let's assume you can walk through walls. However, at the very least you'll want to walk on the floor. What about split levels and stairs?
No. That sounds too hard. We'll rule out walk-through. We've now got a floating camera, and it'll happily pass through any geometry.
Now, should the camera orbit around a point, or is a first-person flying camera more appropriate?
Wait, how fast should the camera even move when you indicate you wish to move forward? Okay, let's make that configurable.
What if you have two points of interest and a large gap in between? How do we get from one area to other? Should the camera travel faster in this area? Should there be some standard mechanism to jump to another location in the scene?
Hmm, too hard. Let's ignore that problem for now.
Should there be constraints with respect to how far you can zoom in and out? I guess we could define that in the file somehow.
Okay.
Animation tracks. They were mentioned. Do they all auto-play? Is that going to be performant? What if we want to trigger animation tracks based on camera location or some other event?
We'll need, like a DOM, I guess?
> while we are not proposing a DOM for the data at the moment, we expect to in the future
Ah, right. Well. Let's assume in the future there's a DOM.
Wait? How do we interact with the DOM? JavaScript somewhere? Where are we collecting user events from to drive this interaction? Can we display UIs within the 3D scene and receive events? How do we build those UIs?
This snowballs real fast. The solution you're going to end up with is WebGL/WebGPU. Anything less than that is far too restrictive and going to make for a terrible user experience.
Re: Apple proposes adding "model" element to HTML that displays 3D content
#114The naming is unfortunate. But also, it'd be pretty cool if this was implemented as a browser extension, that way it could be much faster to iterate over. I wonder why this is not common practice? Modifying browser internals is such a pain that it strikes me as odd that it is the 1 established way of going about this proposals.
If I am not wrong, a functional version can be written in plain JS (+ WASM?) using Custom Elements (tho tag names must contain a hyphen). Except I don't know why all the non-Google stakeholders are hell bent on sabotaging Custom Elements. While Google is doing so much innovation in that space. Strange.
Came here to say that. They could do a custom element implementation, perhaps ship that (in their own namespace), offer polyfills for other browsers and go on from there.
Re: Apple proposes adding "model" element to HTML that displays 3D content
#115My initial reaction to this was "just use WebGL", but they've made a pretty compelling case as to why that's not ideal: > Consider a browser or web-view being displayed in Augmented Reality. The developer wants to show a 3D model in the page. In order for the model to look accurate, it must be rendered from the viewpoint of the user - otherwise it is a flat rendering of a three-dimensional image with incorrect perspe…
> Consider a browser or web-view being displayed in Augmented Reality
Feels like Apple is working on something that needs this.
Re: Apple proposes adding "model" element to HTML that displays 3D content
#116I don't understand why a 3D geometry model would best be called simply "model" in the context of HTML. The term "model" is a pretty generic term that's very relevant to HTML, already used heavily in Web frameworks, etc.
At least they could use a proper name instead of the generic "model", something like "object3d".
They could just use .
Re: Apple proposes adding "model" element to HTML that displays 3D content
#117Interesting use of the example image, I wonder if Apple is thinking of iMessage on the web, after we've seen FaceTime for the browser. I wouldn't have thought it a possibility, but it's strange they used that example.
> I wonder if Apple is thinking of iMessage on the web, after we've seen FaceTime for the browser
My first thought was that Apple is indeed working on a VR and/or AR headset.
Re: Apple proposes adding "model" element to HTML that displays 3D content
#118My initial reaction to this was "just use WebGL", but they've made a pretty compelling case as to why that's not ideal: > Consider a browser or web-view being displayed in Augmented Reality. The developer wants to show a 3D model in the page. In order for the model to look accurate, it must be rendered from the viewpoint of the user - otherwise it is a flat rendering of a three-dimensional image with incorrect perspe…
I don't understand why passing a new view matrix to webGL is such a security risk or why that needs to be shared with he site owner.
I also think a window with a dynamic view matrix would make more sense in certain cases. Some techniques work when rendered to a surface that would not work in VR, ie screen space effects.
Re: Apple proposes adding "model" element to HTML that displays 3D content
#119Based on the comments so far, I think I'm in the minority who think this would be a good feature for the web. This proposal might not be perfect, but I think the idea is great. There's lots of good reasons to semantically want to embed a 3d model in a page, like any other type of media. Say you're writing a course explaining how differential gears work - a 3d model of that could be really handy to look at to learn. A…
Years of trying to treat models like images shows us it's a poor idea. Maybe now with Disney/Epic PBR being the standard material model it's closer but, that Panic truck? Good luck defining the exact lighting for that so it renders correctly and consistently. The spec text even says that the lighting environment will come at a later date. So, no, you don't seemingly have any control over one of the most important par…
>This proposal does not aim to define a mechanism that allows the creation of a 3D scene within a browser using declarative primitives or a programmatic API.
the goal isn't to fully render scenes, it's simply to gain the ability to include a 3d resource in the page semantically. heavier solutions to render scenes already exist.
Re: Apple proposes adding "model" element to HTML that displays 3D content
#120Earlier quoted context omitted.
> 3D could be more popular if it was easier to share and display. Could be more popular? The APIs are already there, literally nothing stops you from writing a pure js lib that supports 3d models in a way that you want and can be reused on any site that needs it. Video didn't have that, video needed 3rd party plugins to even run. Video still managed to attract a sizeable user based before that. > You're right, browse…
> Could be more popular? The APIs are already there, literally nothing stops you from writing a pure js lib that supports 3d models in a way that you want and can be reused on any site that needs it. The APIs are there, but not the renderer or a way to interact with it. I'm looking at it as a consumer, not a web developer. I don't want to have to write a library or use someones else's, that's the point. Browsers have…
> I'm looking at it as a consumer, not a web developer
Then why do you care if it is part of the browser or just a library? A consumer doesn't need to know that, certainly wont interact with a model element directly.
> you could argue that could be done just fine with html and javascript instead.
Not at the time the video controls where introduced and now many sites use DRM which would prohibit a pure js implementation unless you want to piss of Netflix, Amazon, etc. .
> Should everything that's not as popular then be dismissed?
There are finite browser developers working with finite time and finite resources. Why spend any of that on something a random web dev. can easily implement as a JS library?