-
Notifications
You must be signed in to change notification settings - Fork 192
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Extend popovertarget to support custom elements #1088
Comments
Could you explain more about what you intend to extend? Is the proposal to allow custom elements to have the My opinion is that |
The main goal is to make this example work:
And if we think focusing on
|
@gregwhitworth Could we add this issue to the agenda for the 9/12 meeting? I don't have permissions to add labels. |
@sanketj done, also updated your privileges so that you should be able to modify the labels |
The Open UI Community Group just discussed The full IRC log of that discussion<bkardell_> masonf: There are two directions for the interest target api - a central one, or 'don't do that' - just provide the lower based API that make each thing possible.. Is it better to have a bunch of lower level APIs or a higher level API? If it is lower level it seems like, harder to get the accessibility right? wdyt?<bkardell_> alexking: I am a web platform developer from Microsoft. The issue is about two things: it seems it has been a pinpoint for developers that they can't use a popover (or invoker) on a custom element... <keithamus> q+ <gregwhitworth> q+ <gregwhitworth> ack keithamus <masonf> q+ <bkardell_> keithamus: the reason we picked buttons is for accessibility - they already have the right interaction model, they are exposed the right way, they are the semantically relevant element - none of that is true for custom elements out of the box <bkardell_> q+ <masonf> q+ scotto <bkardell_> keithamus: if it gets commandfor does it automatically get all oft hose things <gregwhitworth> ack masonf <gregwhitworth> q-- <gregwhitworth> q-- gregwhitworth <keithamus> q+ <masonf> https://github.com/whatwg/html/issues/9110#issuecomment-1516468633 <bkardell_> masonf: there's a couple of threads you linked to about this. my pref would be to solve the more general "how do you make a custom element that works like a regular button" - there is a suggestion from annevk that says you can put an element internals that just does all that. That makes sense to me. I think it would be better to solve generally and then use that here <gregwhitworth> ack bkardell_ <gregwhitworth> ack scotto <sanketj_> q+ <bkardell_> scotto: to echo mason, I understand the use case on custom elements - why use a button when I could use a button that smells like me, my own button. We need it to be a button - exposed as a button... you need to get the a11y semantics and so on. You can use JS right now to make it do all of the things, but you have to do a lot of work to make it accessible. Just allowing this on custom elements without getting the promise that what they <bkardell_> are putting this on is a button means it won't work the majority of the time <bkardell_> scotto: so that is the blocker. It should work on a button, but without that author requirement it is going to do harm and people are going to abuse this in ways that won't work for people <gregwhitworth> ack keithamus <bkardell_> alexking: We can agree with that, it makes sense to me <bkardell_> keithamus: There are broader discussion since we don't have is="..", how do we make it into a button? It's not just buttons tho - those are common, so you wind up talking about it, but there are things like tables and lists where you can't have a custom <tr> or <li> because they can't be parsed that way <bkardell_> keithamus: maybe this is too abstract but maybe we need to look at how to open the parser to how we can interleave them, or maybe make some kind of behavior you can mix in... button has the behavior of my button - they are compositional - we discussed some of these ideas at TPAC last year... the internals, I feel like that's ok, but we're going to wind up with a list of those <gregwhitworth> ack sanketj_ <masonf> q+ <bkardell_> sanketj_: this is fantastic feedback - the element internals thing is definitely something to go back and look at. This whole discussion is about button and then making the custom element act like a button, but is it specifically the point that it is button and button like, or is it... like, hypothetically if you wanted to use input type checkbox as a popover host - are there things you could to do enable that while still supporting all of <bkardell_> the rest? <gregwhitworth> ack masonf <bkardell_> masonf: are you saying it's not just custom elements, could we make even input type checkbox as a thing that does a popover... we talked about this while designing it. Everything else that isn't a button is different, that's why we didn't do that - they would have other issues <bkardell_> masonf: there is one implementer who is against builtin custom elements but I agree that it is the better solution <bkardell_> masonf: some browsers ship this - just one doesn't. <bkardell_> sanketj_: you mean like extends? That too had it's limitation, you wouldn't be able to attach a shadow Dom, it still doesn't work <bkardell_> masonf: I'm not saying it fixes all problems, I'm just saying a more general thing - a way that makes it behave as much like a built in would be ideal |
This is a great conversation and I was sorry to yet again miss - Thursdays are brutal for meetings these days :). In terms of built-ins, that could be an option, but given that there is an implementor who is fundamentally opposed, it's not something that can scale well as a solution. The lack of a shadow root aside, if we had broad agreement from implementors, I would be much more open to this as a direction. While the initial conversation is around In terms of using element internals, both of these use cases (and more) could be solved if there were a way to inherit the behaviors and semantics of a native element as part of that API. It seems like that's what was potentially proposed, and if it's feasible, could be a good way to ensure that something signaling as a button performs as a button. It's possible that could help solve the long overdue issue where custom element buttons can't act as submit buttons: WICG/webcomponents#814.
Ultimately, my growing concern is that while the declarative API's that are being iterated on are helpful for part of the web platform, they leave out a first-class citizen in Web Components, which is the web platforms design for componentization - a central delivery method of most design systems these days. If our goal is to advance the web platform, it seems like we need to keep an eye on how these solutions scale to it's own component model. Without that as a core consideration, it feels as though we're not delivering on a holistic solution. |
adding more details for the TPAC discussion: (a full explainer will be published soon) Goal
Proposed Solution
Behaviors Focus PreventDefault formAssociated
Alternative solutions
Full list of potential values for Derived from
Derived from other form elements:
|
If type took precedent over role, that’d be the opposite of how role is supposed to work. Role should still be allowed to modify the role of the element, but the element should still have all the behaviors / functionality of the type that was declared. |
I would expect that element internals would set the default for the element and not take precedence over declarative syntax. Just as native HTML elements have default semantics which can have precedence taken by role, custom elements have this today (constrained) with ElementInternals. If I define the default role as "button" for my custom element with ElementInternals and then in HTML land add Just a note, but speaking specifically of button, this would solve a very longstanding issue and pain point with form associated buttons in that they cannot be a submit button: WICG/webcomponents#814. Building off the very primitive example provided by Domenic, using elementInternals would take the strawman proposal slightly further but allow reusable, encapsulated, semantically correct, and likely more performant custom elements to be delivered at scale for design systems. In the Webkit position on customized built-in elements, they brought up some very fair concerns about customized built-ins that this proposal may be able to address as well: WebKit/standards-positions#97 (comment) |
@chrisdholt yup that's what i'd expect. reading through this again, maybe i misread how the role/type were taking precedent over each other? but sounds like we agree, so if i did read that wrong, never mind me :) |
yes, to clarify, the order of precedence is |
Pasting in minutes from the ad-hoc extension of the joint WHATWG-CSSWG-OpenUI CG session: Extend popovertarget to support custom elements sanketj: use case we're looking to enable is popovertarget on custom elements sanketj: allow authors using custom element to ?? without any script sanketj: we believe the proposal can be extended to other types of form control behavior sanketj: proposal leverages element internals sanketj: give the cusotm element button-type button behavior, support things like popovertarget/action sanketj: also a11y things, if no role defined sanketj: easy opt-in to communicate to the browser that this custom element should have button-type behavior on the page chrishtr: so if clicked, will open popover? sanketj: behavios just like , so if popover defined chrishtr: so when click bubbles up it would be the default action? sanketj: yes. can be preventDefault'ed lea: what is current behavior? sanketj: nothing. it does nothing annevk: it's only an attribute for buttons sanketj: you'd have to write script to call popover methods in JS lea: are we limiting to custom elements with [limits] or any element? sanketj: you'd have to give it type of button sanketj: today, button regardless of type, can launch popover regardless sanketj: maybe if extended later ??? sanketj: custom element needs to resemble the type of the native element it's emulating lea: [missedq] sanketj: set internals, and there you say it's emulating a button lea: other values? sanketj: none yet, but could do submit or others lea: why not use computed role? annevk: roles don't have API around them. They're only used for reflection to AT. annevk: they're not semantic annevk: all of ARIA is just affecting the accessibility tree; intentionally no other side-effect Domenic, you wanted to talk about disabled="" Domenic: Makes sense. one thing to keep in mind is interaction with disabled Domenic: if form-associated, can be disbled Domenic: so will need to check that the same way sanketj: good point keithamus: lots of subtle differences wrt e.g. activation via spacebar or enter etc. keithamus: working inside a form or outside a form, should have consistency. but might be confusing to developers Domenic: spacebar thing is interesting. would adding this make spacebar cause click events? sanketj: for behaviors, have we discussed in WHATWG? keithamus: We resolved that if it's a form participant, needs to have type=button as content attribute keithamus: and has to be a tag of button (or exception for this case) keithamus: so we should make provisions the same sanketj: so strictly needs to be of type button if form-associated, but if not they could be of type reset or submit as well? sanketj: We want this behavior to also extend to ??? things like commandtarget sanketj: ideally work the same keithamus: lots of small assitive pieces like spacebar that have implicit ???? annevk: is that what we do for form-associated elements? Domenic: web devs add own keyboard/click listeners Domenic: if it is going as far as behavior, then maybe it should do the spacebar->click translation also sanketj: proposal is if you set type=button, and no explicit role set anywhere, then it gets implicit role of button sanketj: if element-internal role, then those take precedence lea: I agree ARIA should not produce side-effects, but what about the other way? lea: would type=button cause ARIA to compute to button? Domenic: different about ??, but should reflect in AT keithamus: we already have the different levels annevk^: do we want that many layers of abstraction? annevk: not reflecting in element internals keithamus: I consider that part of raw computation. Another lookup in the if-chain keithamus: aria-reflected will always be whatever content attribute is annevk: idk if it matters about table annevk: but I think we can just put it on the internals Domenic: element.internals role is to define its default role Domenic: if designing button from scratch, would set it Domenic: now saying you can set type Domenic: but could also set internals Domenic: but C++ elements don't have this third way Domenic: For every built-in element there's C++ dfeault role, and there's content attribute. Now we're adding third one. keithamus: we have third raw computation already Domenic: which element? keithamus: I believe there's code that computes roe annevk: but that's just flattening the role, if multiple are specified in the attribute annevk: but that's still only one level, it's not an extra level Domenic: might not agree on how many levels there are, but question is do we want to add the additional leveL? Domenic: [missed] sanketj: so you're proposing it should change element.internals.role sanketj: if you never set role, and you query element.internals.role then you get button annevk: Should this be on internals? Form-associated is on constructor. lea: it's static lea: my high-level concern was to not have authors declare their intent twice lea: otherwise will end up with wrong roles +1 lea lea: so don't mind how it happens, as long as simple case works by default lea: and allow override in internals Domenic: all agreed, but just question of how annevk: do you remember why form-associated is static? Domenic: for inheritance, which webkit doesn't support Domenic: could do autonomious custom button inheritance keithamus: common practice ot have a base classs and then extend it keithamus: can't think of why you'd have inheritance chain for button annevk: e.g. submit button subclassing from button Domenic: if you call super, then the superclass will set element.internal.type keithamus: a lot of internals can only be captured once, so if you are defining ? Domenic: oh, basically can't use element.internals for inheritance keithamus: element.internals throws on the second call annevk: so we have static form-associated, should this match that existing pattern yes/no bkardell_: I like it, but annevk: Oh, I wanted to say that too bkardell_: but explainer doesn't mention :focus-visible, but that should also apply bkardell_: we did so much work to make :focus-visible work, if we make it so that doesn't work it'll be a step backwards annevk: focus for custom elements is an unsolved problem annevk: don't know which kinds of elmeents to opt in Domenic: problem with focus is, say "want to be keyboard focusable or mouse focusable" but what you actually want is to match platform conventions Domenic: so best proposal was "behave like a button" Domenic: tried to be too generic Domenic: but this is more specific, so can solve it annevk: I guess that works Domenic: not 100% certain it'll decompose well with the hypothetical future, but 80% sure annevk: if just going to be default buttons or submit buttons? Domenic: I hope they have same focus behavior annevk: yes but different selector matching sanketj: intent is type=button is button type=button sanketj: if you want to make it submit, then need type=submit <bkardell_> selector matching is a good point too annevk: Maybe do more than just type? or is it open to e.g. hyperlink, submit button, etc. Domenic: yes that's the idea annevk: so really big enum keithamus: wrt roles, image has non or presentation depending on herusitics keithamus: table has ?? Domenic: those are set by a single property, not two layers keithamus: Not to derail too much, but what if HTML button had a superclass that was designed for this kind of inheritance pattern keithamus: woudl be like "is" but sidestepping issues around custom eleents keithamus: would that ... work? sanketj: one thing we want to ensure, challenge today, is attaching shadow DOM sanketj: can't do for something that inherits from HTMLButtonElement keithamus: I'm proposing a new class, e.g. HTMLAbstractButton, between HTMLElement and HTMLButtonElement ??: do we have anything like that today? annevk: nope annevk: but none of those can be extended and used as custom element keithamus: could we extend HTMLHeadingElement? annevk: I'd be worried about changing inheritance of existing elements Domenic: why is this better than element.internals type? keithamus: I think it's more consistent for developers keithamus: authors want to extend more than just button keithamus: we're already talking about other types of buttons and inputs keithamus: not necessarily a point against this proposal, but ... Domenic: both would work annevk: either way would need to go through element.internals annevk: so not getting much except the type annevk: and it gives us longer prototype changes, and compat headaches annevk: maybe impact is negligible but doesn't seem great? the superclass idea is really interesting annevk: if there was a concrete benefit Domenic: conflicts with any base classes that authors might want to use, because can't have more than one acl leaverou lea: Value space of this new property... looking at thread seems to mix input types with element types leaverou, you wanted to ask about property values lea: e.g. range -> input type=range; textarea -> textarea lea: worried it might not be future-proof lea: perhaps a pseudo-selector syntax might work? e.g. writing out the tag name and if want to specify similar to input type=range use input[type=range] annevk: could we use input-range? lea: introduces a new microsyntax Domenic: doing that anyway, but it's a subset of an existing one annevk: but enums are much easier to parse lea: wouldn't need a full CSS parser sanketj: type as an attribute is very familiar, in input and button sanketj: so keeping alignment would be good for developers input[type=range] vs input[type="range"] would be surprising if only one of them work Domenic: I'm sympathetic with this worry. We had this problem when solving focus thing Domenic: didn't come up with a good solution Domenic: but also this proposal is very simple, and that's attractive Domenic: do we have anticipate conflicts between tag names and input types? Domenic: I could see that happening (total side comment on an earlier discussion: the details of the keyboard activation behavior for buttons are awful, see https://issues.chromium.org/issues/40863985#comment6 ) lea: we already have conflicts, like search Domenic: there are complexities to attribute selector syntax, but also sympathetic because devs might find it useful lea: also room for future extension, e.g. if we need to identify children of a particular eleents annevk: it's not a microsyntax, it's an enum annevk: what if you put random other things in? ??: could restrict it to [missed] lea: just restrict it to '['... zcorpan: it's surprising to have a syntax that looks like a selector but parses differently zcorpan: e.g. people will expect to be able to use quotes here zcorpan: also, select element has two different kinds of widgets, which have a complex way of deciding which one to use zcorpan: so select[multiple] or select[size] lea: if we went with enum approach, how would you distinguisH? zcorpan: select-dropdown vs select-somethingelse for example lea: but now moved to unfamiliar keywords lea: introduces more API surface lea: if connecting to something already familiar with, then easier to relate annevk: if know about select, select-optionA and select-optionB is not that hard (if it's select-single and select-multiple?) lea: how would it grow in 10 years? Domenic, anne: probably only ~ 5 values lea: ideally this should serve all the "behaves like this element" use cases keithamus: value of button adheres to both options, so let's resolve on button and we can have this argument when we need it? lea: I think 2 main classes of use cases lea: custom elements that behave like buttons lea: and custom elements that wrap buttons lea: in latter case, which element do you want to actually be the popover target? lea: whereas in first case you generally want to make that custom elemet a popover target lea: Seen it come up many times, a big class of custom elements is simply wrapping a native element and adding some scaffolding / styling lea: essentially want to forward things to the shadow DOM built-in element lea: what do people think? ???: for subset of that problem, doing a breakout tomorrow lea: worried we solve this as smaller problems and not seeing the forest for the trees gregwhitworth: scenario in which you bind button via popover target to custom element ... is it because you don't know the ID? why is it a problem sanketj: wrapped button is largely to get around the limitation that we are trying to solve sanketj: so if we solve this, maybe we'll start seeing less of that pattern bkardell_: was going to say the same bkardell_: we've talked about this kind of thing before bkardell_: e.g. referencetarget @fantasai @sanketj I don’t think so, happy to discuss further but don't want to derail this discussion bkardell_: some of it comes out of Apple's objection to 'is' bkardell_: is it a button, or does it contain a button? Is it compositiion or inheritance? keithamus: one way to get around lack of built-in is to use referencetarget lea: still need an easy way to forward internals in an API lea: right now need to reimplement everything bkardell_: I think for button this works fine, but it's the easiest bkardell_: how realistic is it to extend bkardell_: do you run into the same kind of problems if you'r doing inheritance-based things for e.g. select bkardell_: how can you say "this is a select" bkardell_: wondering if we've thought about those limits, are we ok with them bkardell_: will we end up with multiple different solutions bkardell_: for button we can infer so much of the magic, we know exactly what it means to be a button bkardell_: no parts to it sanketj: type attribute, when we add things to it, have ability to think through -- does it make sense? It's not allowing eveyrthing annevk: and what additional API params you need for it bkardell_: it took so long to get one... annevk: good to start. let's start with button! @domenic btw that proposal also uses "meta-selectors" to specify a class of elements, whose grammar is defined exactly how I described :) Indeed, there was a reason we didn't ship it :) sanketj: seems to be overall support for working on this sanketj: some issues to note, e.g how does type reflect into role; how to deal with element/attribute name conflicts sanketj: we'll work on it annevk: can you open issue agianst HTML spec also? sanketj: yes bkardell_: would you pick one more besides button to think through? sanketj: currently submit/reset sanketj: but label could be another one annevk: those are often requested (other than reset) chrishtr: can we resolve on what we agree on so far? PROPOSED: elementInternals.type='button' enables custom elements to get button-like semantics https://open-ui.org/components/richer-text-fields.explainer/ RESOLUTION: elementInternals.type='button' enables custom elements to get button-like semantics |
Can we get a declarative mechanism for |
I think we would need declarative custom elements to unlock those benefits. Right? |
We (Edge) are interested in exploring how we can extend popover API to support custom elements.
Two questions:
popovertarget
to support custom elements? The extension has been discussed in several forums (eg., #9110, #9109 in whatwg, and #302 in openui), so far no conclusion has been reached yet, but from what I can find, it seem the community in general believes that the API would be cleaner and easier to use if we could use popover/popovertarget declaratively cross root.@mfreed7 @keithamus @sanketj
The text was updated successfully, but these errors were encountered: