On Wednesday, November 29, 2023 at 2:23:12 PM UTC+9 Yoav Weiss wrote:
I ran into a Chromium-internal case where having a `close` signal would've helped me reduce complexity significantly. So I'm very excited about this work. I also believe <https://bugs.chromium.org/p/chromium/issues/detail?id=1487835> there's an equivalent mechanism in Electron. Might be worthwhile to get those folks involved and see if there are any learnings to be had. On Tue, Nov 28, 2023 at 9:34 PM Torne (Richard Coles) <to...@chromium.org> wrote: Android WebView exposes the ability to create and use MessagePorts to the host app through a Java API, and currently one of the limitations of this API is the lack of a signal about the channel being closed, which in some cases can lead to apps holding on to Java objects longer than intended. So, I'm very happy to see this proposed for the web API side, as this should hopefully also enable us to support the same thing in WebView's Java API. +Peter Birk Pakkenberg <pb...@google.com> has been tackling some issues here recently. On Tue, 28 Nov 2023 at 12:31, Nonoka Muraki <murakinon...@chromium.org> wrote: Contact emails murakinon...@chromium.org, m...@chromium.org, rak...@chromium.org Explainer https://github.com/fergald/explainer-messageport-close <https://github.com/fergald/explainer-messageport-close#exposing-that-cross-origin-navigation-has-occurred> Specification https://github.com/whatwg/html/pull/9933 Design docs https://docs.google.com/document/d/1lXZU2Pk2ycitqj8aL9kxT2aauwXqp A0vJDUalkXA29Y Summary Notifies one of the MessagePorts that the other port has been disentangled (i.e., close() is called, an owning document is destroyed, or the port is garbage collected). Blink component Blink <https://bugs.chromium.org/p/chromium/issues/list?q=component:Blink> Motivation Channel Messaging API enables two separate scripts running in different browsing contexts to communicate directly. Once MessageChannel is created, we can access two entangled MessagePorts (MessageChannel.port1 and MessageChannel.port2), transfer port2 to the other browsing context, allowing these two scripts to post and receive messages. The problem with this API is that there is currently no timely and reliable way to detect when a MessagePort becomes disentangled. This makes it difficult to release resources associated with ports. Some users have used the unload event to detect disconnection, but it is unreliable especially on mobile. Therefore, the approach available now is to hold a MessagePort in WeakRef and periodically check if this port has been garbage collected by using deref() or to make use of the pagehide event. However, this event is also not an effective solution. The reasons are as follows: 1) It doesn't work when the document crashes. 2) The port must cooperate in signaling the port is going to be disentangled.(i.e., the port must post the message which is "it is going to be closed" to the entangled port.) 3) It fires even if the page is not being destroyed and just enters BFCache. Adding a close event helps to detect timely and reliably that MessagePort is disentangled and make it easier to release resources associated with ports. Initial public proposal https://github.com/whatwg/html/issues/1766#issuecomment-1708027883 Search tags close <https://chromestatus.com/features#tags:close>, onclose <https://chromestatus.com/features#tags:onclose> TAG review Not needed because This is a small feature where we just dispatch a new event. Unfortunately that's not a criteria for skipping a TAG review. Can you file one? I'm concerned by this because every TAG review I've seen in the last couple of years has taken months to get a response. If our own privacy review is positive and we have agreement with other vendors would we block on the TAG review? F Risks Interoperability and Compatibility None Gecko: Positive (https://github.com/mozilla/standards-positions/issues/914 ) There are definitely positive signals on that thread, but it's not yet labeled as "positive" AFAICT. Mozilla people are already involved in the initial Github issue <https://github.com/whatwg/html/issues/1766>. WebKit: No response (https://github.com/WebKit/ standards-positions/issues/271) WebKit people are already involved in the initial Github issue <https://github.com/whatwg/html/issues/1766>. Web developers: Positive (https://github.com/whatwg/html/issues/1766) Other signals: WebView application risks Does this intent deprecate or change behavior of existing APIs, such that it has potentially high risk for Android WebView-based applications? None Debuggability None Is this feature fully tested by web-platform-tests <https://chromium.googlesource.com/chromium/src/+/main/docs/testing/web_platform_tests.md> ? Yes. https://chromium-review.googlesource.com/c/chromium/src/+/5003089 Flag name on chrome://flags None Finch feature name None Non-finch justification This is a small feature (just dispatching a new event) that doesn’t need experimentation. Requires code in //chrome? False Tracking bug https://bugs.chromium.org/p/chromium/issues/detail?id=1495616 Estimated milestones No milestones specified Link to entry on the Chrome Platform Status https://chromestatus.com/feature/5086799659794432 This intent message was generated by Chrome Platform Status <https://chromestatus.com/>. -- You received this message because you are subscribed to the Google Groups "blink-dev" group. To unsubscribe from this group and stop receiving emails from it, send an email to blink-dev+unsubscr...@chromium.org. To view this discussion on the web visit https://groups.google.com/a/ chromium.org/d/msgid/blink-dev/CAA66OT2mWU9WJnkVzD2gm_%2Bk%2Bewag5Lhq_ K39Vo9WGdtGW79cQ%40mail.gmail.com <https://groups.google.com/a/chromium.org/d/msgid/blink-dev/CAA66OT2mWU9WJnkVzD2gm_%2Bk%2Bewag5Lhq_K39Vo9WGdtGW79cQ%40mail.gmail.com?utm_medium=email&utm_source=footer> . -- You received this message because you are subscribed to the Google Groups "blink-dev" group. To unsubscribe from this group and stop receiving emails from it, send an email to blink-dev+unsubscr...@chromium.org. To view this discussion on the web visit https://groups.google.com/a/ chromium.org/d/msgid/blink-dev/CAEV-rjcZL5YnizhWf0xA% 3DBaSt4EmY32p1tjhYG%3DfEOP1B-Kk%2BQ%40mail.gmail.com <https://groups.google.com/a/chromium.org/d/msgid/blink-dev/CAEV-rjcZL5YnizhWf0xA%3DBaSt4EmY32p1tjhYG%3DfEOP1B-Kk%2BQ%40mail.gmail.com?utm_medium=email&utm_source=footer> . -- You received this message because you are subscribed to the Google Groups "blink-dev" group. To unsubscribe from this group and stop receiving emails from it, send an email to blink-dev+unsubscr...@chromium.org. To view this discussion on the web visit https://groups.google.com/a/chromium.org/d/msgid/blink-dev/5b5236d9-f139-4cef-8a09-1a3dd5b51d0an%40chromium.org.
