219005 2020-11-16 14:39:26 -0800 Implement ResizeObserver's devicePixelContentBox 2026-03-12 21:11:50 -0700 1 1 1 Unclassified WebKit DOM WebKit Nightly Build Unspecified Unspecified NEW https://bugs.webkit.org/show_bug.cgi?id=230242 https://bugs.webkit.org/show_bug.cgi?id=291310 InRadar P2 Enhancement --- 157743 244141 1 kbr ahmad.saleem792 apolostudioapps cathiechen cdumez daxpedda dino fred.wang gman jdarpinian jujjyl karl+webkit kkinnunen koivisto luiscastro193 ricardovalero95 rik simon.fraser smoley thorton webkit-bug-importer oldest_to_newest 1708001 0 kbr 2020-11-16 14:39:26 -0800 ResizeObserver has gained the ability to observe the number of device pixels that a given element covers in the "device-pixel-content-box" observation: https://www.w3.org/TR/resize-observer/#dom-resizeobserverboxoptions-device-pixel-content-box This feature was discussed and specified in the CSS working group. Some concerns were raised by WebKit engineers at the time that WebKit can only provide an accurate answer for this observation after, and not before, the paint cycle runs. Even with this concern, this feature is important for web applications that expect to be able to do pixel-accurate rendering. Now that WebKit has ResizeObserver support (thanks to Igalia folks for implementing it!), would it be possible to do at least a first-cut implementation of this new observation? It should make the example in the article https://web.dev/device-pixel-content-box/ render more stably. Currently it shows bad moiré effects in Safari Technology Preview on Retina display MacBook Pros. 1708083 1 webkit-bug-importer 2020-11-16 18:54:03 -0800 <rdar://problem/71469186> 1793394 2 simon.fraser 2021-09-14 12:58:15 -0700 Some discussion that might result in spec changes here: https://github.com/w3c/csswg-drafts/issues/6608 1993886 3 karlcow 2023-11-20 06:08:59 -0800 *** Bug 264158 has been marked as a duplicate of this bug. *** 2117586 4 apolostudioapps 2025-05-16 05:08:30 -0700 any news on this?? 2168376 5 jujjyl 2025-12-27 16:41:53 -0800 A common problem with Unity3D rendered content on Safari is that the render target does not match the browser 100%, since Safari does not implement ResizeObserver with devicePixelContentBox. Any thoughts on Safari to implement this feature? The code pattern that Unity3D utilizes is practically demonstrated at https://webgpu.github.io/webgpu-samples/?sample=resizeObserverHDDPI - which moires on Safari due to lack of web page content being unable to guarantee that canvas contents get 1:1 displayed on screen. 2189870 6 ahmad.saleem792 2026-03-12 21:11:50 -0700 Pull request: https://github.com/WebKit/WebKit/pull/60534