Open UI

by alexzeitler • 2 years ago

20 6 2 years ago

aphroz• 2 years ago

The idea of allowing to easily define the behaviour of the standard UI component is interesting and will save us from having to reinvent the wheel everytime. But why discussing this kind of things on Discord? Is open the new close? Not so good naming if you ask me, it might confuse people.

gregwhitworth3• 2 years ago

With regards to Discord, that's simply used in place of IRC for informal discussions. All potentially "normative" discussions occur within the Github issues and minutes are recorded as with other W3C groups via our IRC Channel. Thanks for the feedback :)

jauntywundrkind• 2 years ago

With apologies, a lot of these feel like a grab bag of elements.

But Keith Cirkel's Invoker/Actions feels like a big paradigm idea, something that de-javascriptizes & encodes in html the primary intents of the web. There's something super enticing here, a slim/trim layer of paradigm, which honestly we have seen little of expressed in the web itself (whereas libraries have spiraled towards their own meaning). (Maybe css scopes being a good counter example forwards.) https://open-ui.org/components/invokers.explainer

(There's also Keith's Interest Invokers, https://open-ui.org/components/interest-invokers.explainer/)

Keith is Github's Catalyst framework's primary author. Very Stencil.js inspired, and a clear predecessor of the ideas here; actions and targets. A similar casting of intent out of js and into the DOM. Declare the world! https://github.com/github/catalyst

Catalyst really had me for a bit, and supposedly had pretty heavy rotation within GitHub as the tool that minted the components we saw. Visible progress has really slowed though, as there haven't seen signs of life towards the expected 2.0. I hope we see a return. Keith's practical frameworking was a great guide that pointed toward these abstract understandings, and I feel like the case for this open-uo proposal would be much stronger if Catalyst could be the lives experience here. If anything could be the lives experience.

gregwhitworth3• 2 years ago

Yeah, all are in various stages. And yeah, Keith's proposal takes the popovertarget to a generic approach and will be great to have in the platform!

Other primitives that we're actively working on in addition to that is focusgroup which I think likewise will be of value.

And finally, <select> - that is the one that is "closest" to getting a full makeover in HTML. We're wanting to fully land that one which will hopefully create a fly-wheel and help "speed" this up a bit.

singularity2001• 2 years ago

Using SwiftUI I recently thought about whether HTML could need a reset with all the clutter removed. This could be it (?) LHTM light hyper text markup?

gregwhitworth3• 2 years ago

100% agreed. Same thought I had when I worked with other UI toolkits; you bring it in and it "requires" a Label by default and thus is accessible out of the box. Swift is a great example.

I doubt we will ever get HTML leaner unless we do a DOCTYPE versioning approach that rips out legacy aspects within the engine but that would enable backwards compat to still exist so that the Space Jam site from the 90 can still run in 2300.

Open UI

Related Stories