June 2021 - webkit-dev - lists.webkit.org

Request for position: CSS scrollbar-gutter
by Felipe Erias 15 Jul '21

15 Jul '21

Hi webkit-dev, This is a request for WebKit's position on the CSS "scrollbar-gutter" property. The spec status is Working Draft. This feature is already implemented in Chrome behind a flag. Spec: https://drafts.csswg.org/css-overflow-4/#scrollbar-gutter-property Explainer: https://github.com/felipeerias/scrollbar-gutter-explainer Existing WebKit bug: https://bugs.webkit.org/show_bug.cgi?id=167335 Summary: The scrollbar-gutter property provides control over the presence of scrollbar gutters (the space which may be reserved to display a scrollbar). This gives Web authors more agency over how their layouts interact with the scrollbars provided by the browser, so they can e.g. prevent excessive layout changes as content expands while avoiding unwanted visuals when scrolling isn't needed. Thanks! Best. Felipe

3 7

Request for position: EME MediaKeySession Closed Reason
by Xiaohan Wang (王消寒) 09 Jul '21

09 Jul '21

This is a request for WebKit's position on the "EME MediaKeySession Closed Reason" proposal: *Links* - W3C EME spec issue/explainer: https://github.com/w3c/encrypted-media/issues/473 - W3C EME spec Pull Request: https://github.com/w3c/encrypted-media/pull/487 - Previous Media Working Group meeting discussion on this topic (with Jer Noble): https://lists.w3.org/Archives/Public/public-media-wg/2020Sep/0004.html Summary A MediaKeySessionClosedReason is proposed to indicate the reason for EME MediaKeySession closure, and the `closed` attribute would return a `Promise<MediaKeySessionClosedReason>` instead of the current `Promise<void>`. Motivation The current EME spec says "the CDM may close a session at any point, such as when the session is no longer needed or when system resources are lost". However, there's no way to specify the exact reason for session closure. In some cases, this is part of normal user flow, e.g. user close laptop lid to put the device into sleep mode, where the player should resume playback. In some other cases, this is due to some unrecoverable fatal error. This proposal updates the `closed` attribute on MediaKeySession to provide a MediaKeySessionClosedReason so that the JavaScript player can handle session closure differently based on the reasons. *Contact email* xhwang at chromium.org Looking forward to your feedback! Best, Xiaohan

1 3

Request for position: CSS spelling and grammar features
by dazabani 01 Jul '21

01 Jul '21

I’m seeking WebKit’s position on the following features in css-pseudo and css-text-decor: • ::spelling-error highlight pseudo-element [0] • ::grammar-error highlight pseudo-element [1] • text-decoration-line: spelling-error [2] • text-decoration-line: grammar-error [3] While these features span two specs, they might be worth considering together, because the new pseudo-elements depend on the new text- decoration-line values for UA stylesheet support (and to be most useful to authors). Note that there’s a bug with a WIP patch for the pseudo-elements [4]. [0] https://drafts.csswg.org/css-pseudo-4/#selectordef-spelling-error [1] https://drafts.csswg.org/css-pseudo-4/#selectordef-grammar-error [2] https://drafts.csswg.org/css-text-decor-4/#valdef-text-decoration-line-spel… [3] https://drafts.csswg.org/css-text-decor-4/#valdef-text-decoration-line-gram… [4] https://bugs.webkit.org/show_bug.cgi?id=175784 Cheers, Delan Azabani Igalia

2 1

Request for position on scheduler.postTask
by Scott Haseley 23 Jun '21

23 Jun '21

Hello WebKit-dev, I’d like to request WebKit’s official position on scheduler.postTask(), an API for scheduling and controlling prioritized tasks. It is specified in [1] and further described in [2]. We’ve presented the API a few times to the WebPerf WG. One concern raised there by rniwa was the risk of priority inversions, which we’ve explored in [3]. We recently finished running an Origin Trial in Chrome. Airbnb, who we’ve been working closely with, experimented with using the API to improve their site’s performance, which they wrote about recently [4]. [1] Spec: https://wicg.github.io/scheduling-apis/ [2] Explainer: https://github.com/WICG/scheduling-apis/blob/main/explainers/prioritized-po… [3] Priority inversion exploration: https://github.com/WICG/scheduling-apis/blob/main/misc/priority-inversion.md [4] Airbnb blog post: https://medium.com/airbnb-engineering/building-a-faster-web-experience-with… Additional Links: - Preliminary TAG review: https://github.com/w3ctag/design-reviews/issues/338 - TAG Spec review (under review): https://github.com/w3ctag/design-reviews/issues/647 - Chrome Status entry: https://www.chromestatus.com/feature/6031161734201344 Thank you, Scott

1 0

Request for position on Cascade Layers
by Rune Lillesveen 22 Jun '21

22 Jun '21

Hello webkit-dev, I'm asking for WebKit's position on Cascade Layers drafted in CSS Cascade Level 5. We plan to start prototyping behind a flag in Chromium in Q1. Explainer: https://github.com/oddbird/css-sandbox/blob/main/src/layers/explainer.md Specification URL: https://drafts.csswg.org/css-cascade-5/#layering Request for TAG review: https://github.com/w3ctag/design-reviews/issues/597 Thanks, -- Rune Lillesveen

1 1

FYI: Chrome will block ports 989 and 990
by Adam Rice 18 Jun '21

18 Jun '21

Just letting you know that Chrome is going to block ports 989 (ftps-data) and port 990 (ftps) in version 93. This is to prevent cross-protocol attacks. See https://github.com/whatwg/fetch/pull/1250 for background. I think it would be good for users if WebKit also blocked these ports. There is an issue at https://bugs.webkit.org/show_bug.cgi?id=226971. Thanks, Adam Rice

1 0

Embedded EWS upgrades
by Matt Lewis 17 Jun '21

17 Jun '21

Hello webkit developers, We are currently going through the process of updating the OS and SDKs of the embedded ews builders and testers. We are temporarily going to take iOS ews down while we perform this as fast as possible. As hosts are updated we will be returning them to active status and processing patches. Thanks for your patience, Matt Lewis

1 0

Removing the ENABLE(CSS_SCROLL_SNAP) flag
by Martin Robinson 15 Jun '21

15 Jun '21

Recently, css-scroll-snap was enabled on all ports. The implementation has also been extended to be mostly platform-independent. Since all ports are compiling this code and it's a fairly basic CSS feature should we go ahead and remove the ENABLE(CSS_SCROLL_SNAP) build flag? I think this would greatly simplify the scrolling code. Thanks! --Martin

3 2

Request for position: preferCurrentTab
by Elad Alon 15 Jun '21

15 Jun '21

This is a request for WebKit's position on adding a dictionary member in MediaStreamConstraints called preferCurrentTab. If preferCurrentTab is set to true, the user agent will display the current tab as the most prominent option in the media-picker which getDisplayMedia invokes. Links: - Explainer <https://docs.google.com/document/d/1Og-TUBJQ4SAgReBr7xrd7y-0ZwkjGJMCqBnzDv7…> - Spec <https://eladalon1983.github.io/prefer-current-tab/> - ChromeStatus <https://www.chromestatus.com/guide/edit/5045313003847680>

3 4

WeakHashMap
by Ryosuke Niwa 12 Jun '21

12 Jun '21

Hi all, I've added WeakHashMap to WTF in https://commits.webkit.org/r278803, which allows the use of WeakPtr as keys. Like WeakHashSet, WeakHashMap does not immediately delete the key or the value when the object pointed by WeakPtr goes away (i.e. WeakPtrImpl::m_ptr becomes nullptr). Instead, it relies on rehashing and amortized deletion to remove these entries. This happens either when enough items are inserted or removed such that the underlying HashMap goes through rehashing or when WeakHashMap's entry is accessed (get, find, advancing an iterator, etc...) or mutation (inserting, removing, etc...) more than twice the number of items in the underlying HashMap. Due to this amortized deletion behavior, there is no unbound growth of HashMap even if no attempt to cleanup null WeakPtr is made. However, if WeakHashMap is never accessed or mutated, all WeakPtrImpl as well as the corresponding hash map values will be kept alive. In the situation where this is undesirable (e.g. WeakHashMap could retain a large object as its values or a large number of keys could become stale), call WeakHashMap::nullReferences() to trigger the cleanup manually. - R. Niwa

1 0