- From: Bramus via GitHub <sysbot+gh@w3.org>
- Date: Wed, 11 Dec 2024 14:42:38 +0000
- To: public-css-archive@w3.org
This conversation has stalled, so let’s bring it back to life.
Summary up until now is that `auto` should be put on hold until we get a better idea about what it should do, and that `match-element` is to be used for identity matching. Input is being asked from @nt1m, @fantasai, and @emilio.
---
Personally I would like to _not_ use `match-element` to generate an internal identifier _(unobservable by authors)_ but assign that behavior to the `auto` keyword.
This for two reasons:
1. You can get the originally requested auto-behavior (the one that first tries the `id` attribute, falling back to an internally generated identifier) using `attr()`:
```css
view-transition-name: attr(id type(<ident>), auto);
```
Yes, it’s a bit of extra typing but:
- It works with attributes other than `id` too
- Authors are being very explicit about what they want
- It can always be backported into a keyword _(or function preferably, see “feature detection” point below)_
2. Feature detection
If you feature detect `view-transition-name: match-element` with `@supports`/`CSS.supports` it currently returns `true`.
```css
@supports (view-transition-name: match-element) {
.item {
view-transition-name: match-element;
}
}
```
The snippet above will pass in both Chrome and Safari today already _(uhoh!)_, resulting in multiple elements getting a `view-transition-name` set to `match-element` which will prevent the View Transition from running.
```
Unexpected duplicate view-transition-name: match-element
<div class="item" id="item-1">…</div>
<div class="item" id="item-2">…</div>
<div class="item" id="item-3">…</div>
<div class="item" id="item-4">…</div>
<div class="item" id="item-5">…</div>
```
As a result an author can not safely feature detect support for `match-element`.
When feature detecting `view-transition-name: auto` it return `false` (*), which would allow author to feature-detect `auto` having some magic behavior instead of it being the actual `view-transition-name` that gets assigned.
(*) Unfortunately only in Safari. Chrome returns `true` here, which would need to get fixed – see https://crbug.com/383409588 which I filed just now.
(Alternatively changing `match-element` to become a function (i.e `match-element()` would also fix this feature detection issue)
--
GitHub Notification of comment by bramus
Please view or discuss this issue at https://github.com/w3c/csswg-drafts/issues/10995#issuecomment-2536191408 using your GitHub account
--
Sent via github-notify-ml as configured in https://github.com/w3c/github-notify-ml-config
Received on Wednesday, 11 December 2024 14:42:38 UTC