Re: [blink-dev] Intent to Ship: Direct Sockets API

Just a quick heads-up: I'm planning to prototype Direct Sockets support in Shared & Service Workers in Isolated Web Apps. You can find more details in the corresponding ChromeStatus entry here: https://chromestatus.com/feature/5169039028256768. There will also be a separate Intent-to-Ship submi

Re: [blink-dev] Intent to Ship: Direct Sockets API

LGTM3 for the IWAs in its current form where it is limited to enterprise customers. This and features similar to this would need to be revisited if IWAs shift to a more public audience. If and when that happens, it would be nice to understand the vetting / review process to judge whether some of th

Re: [blink-dev] Intent to Ship: Direct Sockets API

The Promise returned by the "opened" attribute serves that purpose. We used this pattern because it aligns with the WebSocketStream and WebTransport constructors. Reilly Grant | Software Engineer | reil...@chromium.org | Google Chrome On Wed, Aug 28, 2024 at 8:55 A

Re: [blink-dev] Intent to Ship: Direct Sockets API

On Mon, Aug 19, 2024 at 6:46 AM 'Randell Jesup' via blink-dev < blink-dev@chromium.org> wrote: > This was not closed by Mozilla with no opinion, it was closed as Harmful. > Hi Randell! I suspect this is marked as Closed by the tool because the issue

Re: [blink-dev] Intent to Ship: Direct Sockets API

In today's API OWNERS meeting, Jeff raised the concern that this API design doesn't follow our general guidance for having a Promise-returning `request*()` method that integrates with the Permissions API and allows the UA to inject controls (including, potentially, user prompts). This is our lo

Re: [blink-dev] Intent to Ship: Direct Sockets API

LGTM2 On 8/28/24 8:03 AM, Yoav Weiss (@Shopify) wrote: LGTM1 - the use case is clear, as well as the dangers of enabling this over the open web. Confining this API to IWAs seems to strike a reasonable balance. On Mon, Aug 19, 2024 at 3:46 PM 'Randell Jesup' via blink-dev wrote: This w

Re: [blink-dev] Intent to Ship: Direct Sockets API

LGTM1 - the use case is clear, as well as the dangers of enabling this over the open web. Confining this API to IWAs seems to strike a reasonable balance. On Mon, Aug 19, 2024 at 3:46 PM 'Randell Jesup' via blink-dev < blink-dev@chromium.org> wrote: > This was not closed by Mozilla with no opinio

Re: [blink-dev] Intent to Ship: Direct Sockets API

This was not closed by Mozilla with no opinion, it was closed as Harmful. Randell Jesup, Mozilla Networking Team On Tue, Aug 13, 2024, 2:59 PM Chromestatus wrote: > Contact emails greengr...@google.com > > Explainer > https://github.com/WICG/direct-sockets/blob/main/docs/explainer.md > > Specif

Re: [blink-dev] Intent to Ship: Direct Sockets API

LGTM as an IWA OWNER (3x LGTM from Blink API OWNERS are still required according to the IWA-specific API launch process ). This API is an excellent example of a case where a small number of applications need a powerful capabilit

Re: [blink-dev] Intent to Ship: Direct Sockets API

+ iwa-team@ for visibility. On Tuesday, August 13, 2024 at 7:20:25 PM UTC+2 Christian Biesinger wrote: On Tue, Aug 13, 2024 at 9:59 AM Chromestatus < ad...@cr-status.appspotmail.com> wrote: Contact emails green...@google.com Explainer https://github.com/WICG/direct-sockets/blob/main/docs/expla

Re: [blink-dev] Intent to Ship: Direct Sockets API

On Tue, Aug 13, 2024 at 9:59 AM Chromestatus < ad...@cr-status.appspotmail.com> wrote: > Contact emails greengr...@google.com > > Explainer > https://github.com/WICG/direct-sockets/blob/main/docs/explainer.md > > Specification https://wicg.github.io/direct-sockets > > Summary > > Allows Isolated W