Intent to Deprecate and Remove: Nonstandard getUserMedia audio constraints

[Guido Urdaneta]

Contact emails

gui...@chromium.org

Explainer

None

Specification

https://w3c.github.io/mediacapture-main/#constrainable-properties 

Summary

Blink supports a number of nonstandard goog-prefixed constraints for getUserMedia from some time before constraints were properly standardized.

Usage has gone down significantly ~0.000001% to 0.0009% (depending on the constraint) and some of them do not even have an effect due to changes in the Chromium audio-capture stack. Soon none of them will have any effect due to other upcoming changes in Chromium's audio stack.

Blink component

Blink > MediaStream

Motivation

These goog-prefixed constraints are a relic from the pre-standard getUserMedia times and should not be used by anyone.

Usage is now low enough that they can be removed without causing significant regressions.

Initial public proposal

None

TAG review

None

TAG review status

Not applicable

Risks

Interoperability and Compatibility

The interoperability risk is zero since these goog-prefixed constraints are not implemented by any other browser. This change encourages developers to use standard getUserMedia constraints, which are supported by all major browsers.

There is limited compatibility risk in that these constraints will now be ignored by Chromium if applications try to use them. This is already partially happening in practice because, while the constraints are still exposed on the Web, several of them have no effect on how audio is captured. They are implemented as custom dictionary members which will be ignored if not present and only as part of a nonstandard constraint syntax.

Gecko: No signal

WebKit: No signal

Web developers: No signals

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?

Same as general compatibility risk.

Debuggability

None

Is this feature fully tested by web-platform-tests?

No

Flag name on chrome://flags

None

Finch feature name

None

Non-finch justification

None

Requires code in //chrome?

False

Estimated milestones

Deprecation in 133 and removal in 134.

Link to entry on the Chrome Platform Status

https://chromestatus.com/feature/5097536380207104

[Philipp Hancke]

This is googEchoCancellation2 and friends seen on (and very often copied from) https://webrtchacks.com/hangout-analysis-philipp-hancke/ (we have come a long way since 2014)?

My impression was that developers kept copying those constraints without understanding the implications or even knowing how to validate a constraint has an effect (and as you say, some did not have an effect anymore).

Not going to miss those, 🔥

--
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 visit https://groups.google.com/a/chromium.org/d/msgid/blink-dev/CA%2BBuZxbvN7WLdnJhLW3ZXKjAdrvAOA7%3DtosmKvKO0yB1-9k3Vg%40mail.gmail.com.

[Mike Taylor]

On 11/7/24 8:13 AM, 'Guido Urdaneta' via blink-dev wrote:

Contact emails

gui...@chromium.org

Explainer

None

Specification

https://w3c.github.io/mediacapture-main/#constrainable-properties 

Summary

Blink supports a number of nonstandard goog-prefixed constraints for getUserMedia from some time before constraints were properly standardized.

Usage has gone down significantly ~0.000001% to 0.0009% (depending on the constraint) and some of them do not even have an effect due to changes in the Chromium audio-capture stack. Soon none of them will have any effect due to other upcoming changes in Chromium's audio stack.

Can you provide a list of each of the prefixed constraints w/ a link to UseCounters? It would be helpful to know exactly what we're approving to be removed. :)

[Guido Urdaneta]

Here's a list of the goog-prefixed constraints and their counters:

Constraint	Counter
googEchoCancellation	https://www.chromestatus.com/metrics/feature/timeline/popularity/1947
googEchoCancellation2	https://www.chromestatus.com/metrics/feature/timeline/popularity/1948
googDAEchoCancellation	https://chromestatus.com/metrics/feature/timeline/popularity/1958
googNoiseSupression2	https://www.chromestatus.com/metrics/feature/timeline/popularity/1954
googAudioMirroring	https://www.chromestatus.com/metrics/feature/timeline/popularity/1957
googHighpassFilter	https://www.chromestatus.com/metrics/feature/timeline/popularity/1952
googArrayGeometry	https://chromestatus.com/metrics/feature/timeline/popularity/1956
googBeamForming	https://chromestatus.com/metrics/feature/timeline/popularity/1955
googExperimentalAutoGainControl	https://chromestatus.com/metrics/feature/timeline/popularity/1950
googExperimentalEchoCancellation	https://chromestatus.com/metrics/feature/timeline/popularity/1948
googExperimentalNoiseSuppression	https://chromestatus.com/metrics/feature/timeline/popularity/1954
googNoiseReduction	https://chromestatus.com/metrics/feature/timeline/popularity/1959
googPowerLineFrequency	https://chromestatus.com/metrics/feature/timeline/popularity/1960
googTypingNoiseDetection	https://chromestatus.com/metrics/feature/timeline/popularity/1953

Note that we also want to deprecate and remove googAutoGainControl and googNoiseSuppression. 

The counters for these two have higher values because they also include the counts for the corresponding standard constraints autoGainControl and noiseSupression as they are considered the same at the point where they are being counted.

googAutoGainControl	https://www.chromestatus.com/metrics/feature/timeline/popularity/1949
googNoiseSuppression	https://www.chromestatus.com/metrics/feature/timeline/popularity/1951

They have similar values to (the also standard) echoCancellation (https://chromestatus.com/metrics/feature/timeline/popularity/1930), so we are confident that those counters are expressing standard usage, so the goog-prefixed variant can be removed.

[Chris Harrelson]

LGTM1

Thanks for removing these!

To view this discussion visit https://groups.google.com/a/chromium.org/d/msgid/blink-dev/CA%2BBuZxYxMaA_K2Q1XxdZ95yurBgghVTWUC8VDv%3DmpwJ7L5k1Ow%40mail.gmail.com.

[Mike Taylor]

Thanks for the list - and LGTM2.

[Mike Taylor]

Let me amend that to LGTM2 % filling out and request the Enterprise, Debuggability, and Testing chips in chromestatus

[Chris Harrelson]

Hi, please also fill out the Enterprise, Debuggability and Testing review bits, then we can approve in the tool as well.

To view this discussion visit https://groups.google.com/a/chromium.org/d/msgid/blink-dev/7e84c90f-4d0d-46c2-95e8-d5842b835077%40chromium.org.

[Guido Urdaneta]

Thanks. I've requested the corresponding approvals. 

Sorry I missed that part before sending the intent. I got a bit confused by the tool's "prepare to ship" option when I was looking for a "prepare to deprecate and remove", so I drafted the email manually.

We'll wait for all the approvals in the tool before making any changes.

[Daniel Bratell]

LGTM3

/Daniel

To view this discussion visit https://groups.google.com/a/chromium.org/d/msgid/blink-dev/1dd55043-35e2-4ee7-8ff0-378331b104d1%40chromium.org.