W3C

– DRAFT –
Open UI Telecon

10 February 2022

Attendees

Present
bkardell_, davatron, davatron5000, davidluhr, flackr, iopopesc_, melanierichards, scotto_, stephstimac
Regrets
-
Chair
-
Scribe
gregwhitworth

Meeting minutes

How to use the <selectmenu>

dandclark: *presents presentation* https://docs.google.com/presentation/d/1aeDak7iBqk95r5KwnYM2QOYXNPS1EihsooA67LFG4Mg/edit?usp=sharing

<miriam> The demos: https://microsoftedge.github.io/Demos/selectmenu/

<una> beautiful demos

<masonf> +1 awesome demos!

<davatron5000> These demos are incredible.

masonf: awesome demos

masonf: you don't need to go demo by demo

masonf: how much JS was needed for those

patrick: most are CSS & HTML, but when re-positioning the popup required JS and the multiselect

patrick: otherwise everything was CSS

patrick: almost all JS was positioning and thus anchor-position would help

bkardell_: echo the demos are amazing

bkardell_: whether we could get the demos

bkardell_: I am curious mainly about the accessibility characteristics

bkardell_: there were a couple that I thought to understand how they work

patrick: none at this point

dandclark: the a11y stuff is wired up by default by behavior

dandclark: so it should work across the gradient as long as the behavior's are applied

dandclark: if you're adding non-trivial interaction inside the listbox they'll need to add the aria

davatron5000: great demos

davatron5000: is it the same as select regarding the value?

dandclark: what gets posted to the form is what gets posted to the selected option. It works the same way as the current select

davatron5000: then I see behavior listbox and button

dandclark: it is how our vision will be how the controller code is wired up

<davidluhr> gregwhitworth: Behavior aspect is key piece that missing in accessibility discussion. The behavior attribute used to be referred to as part. Once we standardize on parts, that allows us to concretely say what accessibility attributes will become.

<davidluhr> gregwhitworth: We basically took an MVC approach. The behavior essentially allows you to connect to the Controller code. If the accessibility isn't as expected, please file issues so we can update how this works.

<Zakim> gregwhitworth, you wanted to respond to bkardell_

scotto_: you've done a great job on this

scotto_: regards to some of the a11y stuff, I've mentioned a few of the issues

scotto_: I've found a few issues but I'm sure there are really intteresting path way forward we can take here

scotto_: we want to also do something a little bit more

scotto_: the level of arbitrary content that can be added in to make it a11y out of the boxx

scotto_: there will be issues but I'm excited to solve them

dandclark: I do owe you a response but I agree with you

<patrickbrosset> Demos are available here: https://microsoftedge.github.io/Demos/selectmenu/

<bkardell_> +1 scotto_ said what I was trying to articulate much better.

<patrickbrosset> And docs: https://open-ui.org/prototypes/selectmenu

Popup proposal

<masonf> presentation: https://docs.google.com/presentation/d/1J9iShu99GRWYKGYVxRq3YrdaVMtE_8HyzRxey5my984

masonf: *presents slides posted above*

<masonf> explainer: https://github.com/mfreed7/popup

davidluhr: I really like this new direction

davidluhr: it allows us to author very tailored experiences

davidluhr: only question was the hidden attribute use it to toggle

davidluhr: details/summary and dialogue utilized the open attribute

davidluhr: have you considered that?

masonf: we can discuss that

masonf: hidden is already wired up so that's why I went that way. Hidden I think has issues animating too

bkardell_: I have some minor things to think about

bkardell_: I *really* like the approach you're taking here

bkardell_: we've discussed this for years and it's clearly a very useful thing

bkardell_: that actually is used in IDL

bkardell_: this would become a global attribute mixin, should we limit it somehow?

bkardell_: the other question is how does positioning work?

masonf: it feels like it should be limited

masonf: there are probably some other limits that should be applied

masonf: we'd like to be as general as possible

masonf: positioning, yes we need to discuss this which is the anchor-position proposal but isn't exactly this proposal

<davidluhr> The `triggerpopup` attribute is a great idea as it allows for non-JS implementations and the triggering button(s) to be located anywhere in the DOM without being limited by a wrapper element.

bkardell_: it's hard to think about without something there

bkardell_: you could create them easily but without you saying how you position them I don't know

<davidluhr> gregwhitworth: Love this. I have a tactical question.

<davidluhr> gregwhitworth: It sounds like there's general interest in path forward. Some of the things Brian raised leans toward moving the spec over to Open UI and incubating the spec?

<davidluhr> masonf: Certainly interested in doing that once we're close to consensus about what it should like.

scotto_: thank you very much

scotto_: this is really interesting and just wanted to say that

scotto_: I also was thinking about open vs hidden

scotto_: obviously we've talked about that

scotto_: one benefit of hidden is you can change the display of that

scotto_: the popup attribute could likewise have a UA mapping to hidden and the open could be used to display it

scotto_: really good feedback we got back here

scotto_: the way that it's used now is that it's visible so get rid of one of the attr that are needed by default

scotto_: to bkardell_ point regarding limiting this, I've thought about that too

scotto_: there is a case where text level semantics don't need a popup

scotto_: I think those can be fleshed out a bit more but I'm really happy where this is now

scotto_: it's what people are asking for with the popup element

<davidluhr> +1 to limiting to specific elements. It would be great if the attributes simply didn't do anything if applied to a non-approved element.

<miriam> This is all very exciting, thank you all!

Minutes manually created (not a transcript), formatted by scribe.perl version 185 (Thu Dec 2 18:51:55 2021 UTC).

Diagnostics

Succeeded: s/davatron5000/dandclark

Maybe present: dandclark, masonf, patrick