For anyone who was suddenly terrified that there's a vast array of very specific and weirdly-named CSS properties on the horizon, fear not - this is something entirely different. This is actually a low-level API for drawing things with CSS, and the funny property names are just what the developers have chosen for these experiments. Check out the link at the bottom of each experiment to see the real code, eg. this one for the "avatar polygons": https://css-houdini.rocks/posts/avatar-polygons/paint.js
I am not a front end developer. Can someone explain the benefit of these low level APIs in a web development context? Is the purpose to eventually have complete GUI toolkits with even the underlying window managers all contained in the browser window? Or is it primarily a performance thing, to get 2d and 3d client side animation? Or something else all together?
Edit: Another possibility -- prettier widgets. I just wonder what need is addressed by making these low level APIs available. (again, not a front end developer, so relatively naive about the whole field, especially state-of-the-art)
It's more like: a lot of web developers have a lot of funky requests we can't possibly all fit in the standard. So instead let's give them the tools to extend CSS themselves.
> "Don’t get me wrong with the title: Houdini is JS-in-CSS (allow JS scripts to be called from CSS). Here, it is more like JS-inside-CSS. You can write the registerPaint’s paint function, directly from CSS. You have access to:"
Being far ahead does not necessarily mean going in the right direction. Considering the implications of some of the web standards I think having a considered and thoughtful approach is probably not wrong.
- https://developer.mozilla.org/en-US/docs/Web/Houdini
- https://drafts.css-houdini.org/
It's not yet widely implemented: https://ishoudinireadyyet.com/