HTML-in-Canvas Demos | Hacker News Reader

HTML-in-Canvas Demos

github.com

67 pointssimonpure28 comments

Tried this demo in Safari: https://arrival.space/htmlcanvas

Looks very cool, and showed a pretty message indicating there's even more:

  HTML IN CANVAS NOT SUPPORTED  
  Use Chrome and enable chrome://flags/#canvas-draw-element

Use Chrome... idontthinkiwill.jpg and aren't we supposed to reject these technologies that allow Google to Embrace, Extend, Extinguish[1]?

Kudos to the artist in spite of this unfortunately esoteric (wish it weren't) concern

[1](hope I'm wrong about it being a triple E https://en.wikipedia.org/wiki/Embrace,_extend,_and_extinguis... )

Seems like you have some pretty strong ranty bias there.

AFAICT, this is a web standard and expected to get buy in from Safari and Firefox before shipping to users. For now it's an experiment you have to specifically enable with flags. No different than any other browser that runs experiments

Here's one from from Apple from 2017

https://webkit.org/blog/7504/webgpu-prototype-and-demos/

Here's another from last year

https://webkit.org/blog/17118/a-step-into-the-spatial-web-th...

> his is a web standard and expected to get buy in from Safari and Firefox before shipping to users.

1. It's not a standard. It's a scribble on a napkin in a working group's repo: https://github.com/WICG/html-in-canvas Created and edited by people from Google.

2. Chrome continuously ships "standards" like this that they create with no buy in and against any and all opposition.

3. Neither of your links have any relation to HTML in Canvas.

> AFAICT, this is a web standard and expected to get buy in from Safari and Firefox before shipping to users.

If it hasn’t already got buy in then it isn’t a web standard, it’s just a Google proposal. Something isn’t automatically a web standard just because Google thinks it’s a good idea.

Here are Mozilla and WebKit positions on this:

> This proposal attempts to solve multiple problems with a single solution. We (Mozilla) recognize the motivation for solving some of the problems, but believe that this is not the right solution to each problem, or in some case a step in the wrong direction.

— https://github.com/mozilla/standards-positions/issues/1076

— https://github.com/WebKit/standards-positions/issues/630

As far as I can see, nobody outside of Google has committed to implementing this.

From the discussion linked in the Webkit repo:

--- start quote ---

Philip: First, google slides is written in svg, so that won't change with this. But google docs is using canvas, so they might be a candidate. … they might want to integrate this peicemeal, this API allows them to start to adopt the feature slowly,

--- end quote ---

This reads to me like "Google Docs decided to go with canvas sometime ago [1], found it to be too hard, so pushed Chrome to have a way to support HTML in Canvas. The rest is just post-hoc justifications"

[1] https://workspaceupdates.googleblog.com/2021/05/Google-Docs-...

What a thread (Mozilla). Thank you much.

> If this ships, this is what we think we'll be facing, but in reverse. Your rendering would become the defacto default. Does this help see where we're coming from?

One side cares about a private, free, open web; the other devs made something COOL and potentially USEFUL (ship it!). Both highly intelligent of course, shockingly different priorities.

The other devs only care about Google and Google's own priorities: https://news.ycombinator.com/item?id=48204981 It also helps that Google is a promotion-package-driven company.

Strongly ranty!

Happy to praise anything good Google does (speedy, reliable YouTube delivery). When they don’t get buy in first, I’m suspicious. They know, but should also care about how bad it is for the web for sites to dictate the browsers we use.

> aren't we supposed to reject these technologies that allow Google to Embrace, Extend, Extinguish

In what way is this technology a means to embrace, extend and extinguish?

This seems like a logical extension of existing web APIs. If we reject everything out of hand then the platform won't improve. It's going through the standards process:

https://github.com/whatwg/html/issues/10650

The standardization process requires implementations before standardization. And the most recent comments on the WHATWG issue are from Jake Archibald (Mozilla) and Anne van Kesteren (Apple). This isn't a unilateral Google project.

Thank you! Did you see the pushback from Mozilla? (downthread - https://news.ycombinator.com/item?id=48208340 )

Given the David & Goliath nature I'd be bending over backwards to avoid situations where I get that kind of pushback, but I may be so utterly out of touch with the reality of how it "has to" be (and then I'm very optimistic that it could still be better).

Maybe this could've been deprioritized until succeeding on other features and learning each other more through that process, then coming to the table with an excellent first draft Mozilla only has polish to add (not fundamental disagreements).

> This isn't a unilateral Google project.

I can list dozens of unilateral Google projects that are pushed under the guise of "standardization" and then released despite no consensus or against explicit objections from other browser vendors.

And this feature? It's literally pushed by Google unilaterally.

HTML-in-Canvas-in-HTML. Yo dawg.

Nah, you forgot the canvas.

I heard you like html-in-canvas demos, but what about canvas-in-html-in-canvas-in-html?

You broke the internet!

Title should have been “html-in-canvas demos in gif on X”

I naively thought the “demo” was a demo, not a X posting by a twit.

non-twit links from page:

https://compiz-web.vercel.app

https://arrival.space/htmlcanvas

Wes isn’t a twit!

I had to poke around a little to learn this:

"The HTML-in-Canvas API lets you draw DOM content directly into a <canvas> or a WebGL and WebGPU texture while keeping the UI interactable, accessible, and hooked up to your favorite browser features."

The word "Canvas" is overloaded in that it can mean dozens of different things. I thought you meant the recently hacked learning management system (LMS) at first.

I remember this repo a few weeks ago: https://github.com/remotion-dev/html-in-canvas

I wonder if it works in more than just Chrome Canary now.

How long until a canvas is used to render the full chrome of a web browser (e.g. including the TLS padlock), showing a fake benign URL in the (fake) address bar while having the user interact with a malicious page?

That's why we have "youtube.com is now full screen" message.

Yes, but this "emergency" UI of the OS could be improved I think. (Also that functionality could have been build easily with normal DOM and JS, cancel and override all events, etc)

Already done, it's called a "browser-in-browser" attack.

Me and another fellow both cited Steve Wittens' review of HTML+dom at each other at the same time a couple days ago. It's a lovely document. https://acko.net/blog/html-is-dead-long-live-html/

At the end he talks about html-in-canvas. This was a while ago, but I think the limitation is still there: the input mapping is totally fudged, just really bad DIY mapping. It only really works for fairly trivial uses.

Steve also mentioned a task I'd like to see: project html onto a curved surface.

I'm so so so happy we're getting something. But it feels like there are many many many constrains here, that might not be so apparent at first glance. It would be so nice to really have HTML in space, but this isn't really there.

Flash is back baby!

In a significantly reduced, dumbed down and non-functional way

Reply on news.ycombinator.com