Intent to Ship: EME persistent-usage-record session

[Xiaohan Wang (王消寒)]

Contact emails

xhw...@chromium.org

Spec

Latest editor’s draft of EME spec: https://w3c.github.io/encrypted-media/

The feature to ship is described at: https://w3c.github.io/encrypted-media/#dom-mediakeysessiontype-persistent-usage-record

Tag review: This feature was discussed extensively during the previous Tag review https://github.com/w3ctag/design-reviews/issues/73, which was closed without concerns (“... we no longer have a concern with persistent usage record”).

Summary

Support a new MediaKeySessionType “persistent-usage-record session”, for which the license and key(s) are not persisted and for which a record of key usage is persisted when the keys available within the session are destroyed. See the spec for more details.

Link to “Intent to Prototype” blink-dev discussion

https://groups.google.com/a/chromium.org/g/blink-dev/c/UcOYfks9jbY/m/zdXzKECzDgAJ

Is this feature supported on all six Blink platforms (Windows, Mac, Linux, Chrome OS, Android, and Android WebView)?

This is an optional feature in the spec. The EME API around persistent-usage-record session will be enabled on all platforms. The actual support by CDMs will vary by platform and by key system supported.

Risks

Interoperability and Compatibility

Interoperability risk: Low. We have received a lot of public support:

• Edge: Supported by previous Microsoft EME spec editor. It is supported by the current owner of the Edge team and the whole flow is already working behind the flag. The equivalent feature is also supported natively by Windows Media Foundation.

• Firefox: No signals

• Safari: Shipped

• Web developers: 

• Proposed and supported by Netflix EME spec editor

• No signals from other web developers

Compatibility risk is the likelihood that a change will break existing web content loaded in Chromium. Since this feature adds a new session type to the Encrypted Media Extensions spec that nobody is currently using, there should be no compatibility risks. Use of this API is entirely optional, so there is no impact on existing applications.

Ergonomics

N/A

Activation

The APIs will be easy to use by developers immediately. However, as most EME features, the license server needs to be updated to support this feature.

Is this feature fully tested by web-platform-tests? Link to test suite results from wpt.fyi.

• https://wpt.fyi/results/encrypted-media, search for persistent-usage-record

• In Chromium, the Clear Key key system doesn’t not support persistent-usage-record session directly, so it’s not possible to fully test it via web-platform-tests. However, there are Chromium browser tests covering the feature using an “External Clear Key” key system which is only used for testing.

Entry on the feature dashboard

https://chromestatus.com/feature/5638708017496064

[Mounir Lamouri]

This is one of the few features hand picked to be added to the EME specification as part of the Media WG charter and was therefore heavily scrutinised. It has cross-browser support and has been pushed by key websites. This LGTM (non-owner)! :)

--
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+...@chromium.org.
To view this discussion on the web visit https://groups.google.com/a/chromium.org/d/msgid/blink-dev/CAF1j9YOhQaSkvS1cLZVvpLw7rXcayJTg2%2BPDoYfKAQPmwV%2BOFg%40mail.gmail.com.

[Manuel Rego Casasnovas]

On 20/07/2020 19:06, Xiaohan Wang (王消寒) wrote:
> Tag review: This feature was discussed extensively during the previous
> Tag review https://github.com/w3ctag/design-reviews/issues/73, which was
> closed without concerns (“... we no longer have a concern with
> persistent usage record”).

There was a comment by dbaron on the tag review in June 11th
(https://github.com/w3ctag/design-reviews/issues/73#issuecomment-642339150)
asking to fill a new TAG review or provide the template information in a
comment. Has anything happen regarding that?

Bye,
Rego

[Daniel Bratell]

In addition to Rego's question, I miss the explainer for this intent.
That makes it harder for everyone to understand the background, purpose
and goal of the change.

/Daniel

[Chris Harrelson]

Also, please use the browser signals format specified in the latest intent-to-ship template, that focuses on official signals from non-Chromium implementations.

--
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+...@chromium.org.

To view this discussion on the web visit https://groups.google.com/a/chromium.org/d/msgid/blink-dev/7ebba63f-409d-af6b-4cdd-d557d5ad63f6%40gmail.com.

[Xiaohan Wang (王消寒)]

Thanks for the comments!

Rego: TAG review template filled as requested: https://github.com/w3ctag/design-reviews/issues/73#issuecomment-663210837

Daniel/Chris: I updated the intent-to-ship with the link to the Explainer doc and new "Signals from other implementations" section.

----------------------------------------------

Contact emails

xhw...@chromium.org

Explainer

https://github.com/xhwang-chromium/encrypted-media/blob/pur-session/persistent-usage-record-session-explainer.md

Spec

Latest editor’s draft of EME spec: https://w3c.github.io/encrypted-media/

The feature to ship is described at: https://w3c.github.io/encrypted-media/#dom-mediakeysessiontype-persistent-usage-record

Tag review: 

This feature was discussed extensively during the previous Tag review https://github.com/w3ctag/design-reviews/issues/73, which was closed without concerns (“... we no longer have a concern with persistent usage record”).

Summary

Support a new MediaKeySessionType “persistent-usage-record session”, for which the license and key(s) are not persisted and for which a record of key usage is persisted when the keys available within the session are destroyed. See the spec for more details.

Link to “Intent to Prototype” blink-dev discussion

https://groups.google.com/a/chromium.org/g/blink-dev/c/UcOYfks9jbY/m/zdXzKECzDgAJ

Is this feature supported on all six Blink platforms (Windows, Mac, Linux, Chrome OS, Android, and Android WebView)?

This is an optional feature in the spec. The EME API around persistent-usage-record session will be enabled on all platforms. The actual support by CDMs will vary by platform and by key system supported.

Risks

Interoperability and Compatibility

Interoperability risk: Low. 

Signals from other implementations (Gecko, WebKit):

• Gecko: No Signal (https://www.w3.org/2020/05/12-mediawg-minutes.html#t07)

• WebKit: Shipped/Shipping (https://webkit.org/blog/10024/release-notes-for-safari-technology-preview-100/)

Web / Framework developers:

• Proposed and supported by Netflix EME spec editor

• The equivalent feature is also supported natively by Windows Media Foundation.

Compatibility risk is the likelihood that a change will break existing web content loaded in Chromium. Since this feature adds a new session type to the Encrypted Media Extensions spec that nobody is currently using, there should be no compatibility risks. Use of this API is entirely optional, so there is no impact on existing applications.

Ergonomics

N/A

Activation

The APIs will be easy to use by developers immediately. However, as most EME features, the license server needs to be updated to support this feature.

Is this feature fully tested by web-platform-tests? Link to test suite results from wpt.fyi.

• https://wpt.fyi/results/encrypted-media, search for persistent-usage-record

• In Chromium, the Clear Key key system doesn’t not support persistent-usage-record session directly, so it’s not possible to fully test it via web-platform-tests. However, there are Chromium browser tests covering the feature using an “External Clear Key” key system which is only used for testing.

Entry on the feature dashboard

https://chromestatus.com/feature/5638708017496064

[Chris Harrelson]

On Thu, Jul 23, 2020 at 1:53 PM Xiaohan Wang (王消寒) <xhw...@chromium.org> wrote:

Thanks for the comments!

Rego: TAG review template filled as requested: https://github.com/w3ctag/design-reviews/issues/73#issuecomment-663210837

Daniel/Chris: I updated the intent-to-ship with the link to the Explainer doc and new "Signals from other implementations" section.

Thanks for doing that. Could you also file a request for standards position here for Mozilla and link to it? It's ok to link to other things and count them as "no signal", but you should also file a request for a standards position, which gives them an opportunity to surface their official view.

[Xiaohan Wang (王消寒)]

Thanks for letting me know. Request for standards position filed here.

[Alex Russell]

Having been part of the previous TAG review for this, I'm happy to see this resolution.

LGTM1

[Xiaohan Wang (王消寒)]

Kindly ping on other owners on this. Thanks!

[Yoav Weiss]

A couple of questions:

* How does this feature interact with e.g. Clear-Site-Data and other browser mechanisms to clear data that sites may have stored?

* Might be a silly question, as I'm not so familiar with EME, but is the persisted information keyed on the top level origin? Or can it be accessed across origins?

Otherwise, the TAG review seems to be in a weird state where it's not clear to me that the folks there know you're asking for a review.

To view this discussion on the web visit https://groups.google.com/a/chromium.org/d/msgid/blink-dev/CAF1j9YMgEq5%2B5iKXRXKHAj93nSXAj49BggtNKJc75iPJv8N_qA%40mail.gmail.com.

[Xiaohan Wang (王消寒)]

Good questions!

Both points are indeed covered in the current EME spec and implemented in Chromium.

• Clear browsing data: The current EME spec requires that "implementations that use Persistent Data must allow the user to clear that data ...". In Chromium, all persistent data stored by the CDM are subject to user clearing as part of the "Clear Browsing Data" process.
• Per-origin storage: The current EME spec also requires to "use origin-specific and browsing profile-specific Key System storage", which is already implemented in Chromium.

For the TAG review, sorry for the confusion. It's up to the Blink process to decide whether we need a TAG review or not, or whether we need a new TAG review if there's already one. Since we already have one which has been reviewed and has concluded, I am reusing that one for this Blink process, and I don't think we need a new TAG review. (Please let me know if you disagree.) It's my bad commenting on the existing TAG review thread, making the owners feel I was requesting a new TAG review, which isn't the case.

Xiaohan

[Yoav Weiss]

LGTM2

[Chris Harrelson]

LGTM3

To view this discussion on the web visit https://groups.google.com/a/chromium.org/d/msgid/blink-dev/CACj%3DBEhsGGfpPOyhKvM0%3DN3SqPezv9gz8VCd64NGggj8LFJhuA%40mail.gmail.com.

[Xiaohan Wang (王消寒)]

Great. Thank you so much for the discussion and approval!

[Chris Harrelson]

Our pleasure. Thanks for your patience.

To view this discussion on the web visit https://groups.google.com/a/chromium.org/d/msgid/blink-dev/CAF1j9YPP-JbRzLVx_Cn4S%3D%2BZ9SnasdG5r74uWP-vy8SM%2BBjE1Q%40mail.gmail.com.