Intent to Remove: MediaStream.stop() method and MediaStream.label attribute

[shiv...@samsung.com]

Contact emails

shiv...@samsung.com

Summary

After adding MediaStream.active attribute, we can remove MediaStream.stop() method and MediaStream.label attribute.

http://w3c.github.io/mediacapture-main/getusermedia.html#idl-def-MediaStream

Motivation

In below link, it has been discussed to remove the MediaStream.stop() method and MediaStream.label attribute,

https://groups.google.com/a/chromium.org/forum/#!msg/blink-dev/m4jiqG67Mvo/j3W-jGLxmQgJ

cr bug: https://code.google.com/p/chromium/issues/detail?id=338500

Compatibility Risk

Low.  since stop() has been implemented on MediaStreamTracks, so the stop method on MediaStream can  be removed.

Usage information from UseCounter

For MediaStream.stop(), https://www.chromestatus.com/metrics/feature/timeline/popularity/537, uses < 0.016
For MediaStream.label, https://www.chromestatus.com/metrics/feature/timeline/popularity/536,  uses < 0.0024

OWP launch tracking bug

https://code.google.com/p/chromium/issues/detail?id=338500

Entry on the feature dashboard

none.

Shivakumar

[Philip Jägenstedt]

Neither MediaStream.stop() or MediaStream.label have been deprecated
yet, so that needs to happen first. It's nice to include the date of
removal in the deprecation message whenever possible, so how many
release cycles do you think is appropriate? Maybe 2?

If you also want to take care of MediaStream.ended at the same time,
that would be great.

> To unsubscribe from this group and stop receiving emails from it, send an
> email to blink-dev+...@chromium.org.

[PhistucK]

​This is an intent to change a prefixed API (webkitMediaStream). What about unprefixing it and simply leaving those out instead?​

☆PhistucK

[Philip Jägenstedt]

Unprefixing is a bigger change that I would obviously support, but
let's not block on that. When unprefixing both interfaces should
simply be aliases, so they can't have different members.

[PhistucK]

They should be aliases if they are really not that different. Encrypted Media Extensions and Media Source Extensions were not aliases, as far as I recall.

Is the Media Stream specification so similar to the current implementation in Blink? If it is, unprefixing is not a big change and can be done now. If it is not, this warrants a non aliased unprefixing.

Does anyone know the current differences between the specification and the implementation?

☆PhistucK

[TAMURA, Kent]

I agree with philipj.  0.016% is not so high, but the usage is increasing.  LGTM if we deprecate them before removing them.

--

TAMURA Kent
Software Engineer, Google

[shiv...@samsung.com]

Thanks, added new patch to deprecate MediaStream.stop() or MediaStream.label link:https://codereview.chromium.org/1194763002/

[Philip Jägenstedt]

In the review, I suggested deprecating these for two release cycles, removing them in M47. If nobody objects that will hopefully land before the M45 branch on Friday.

[owe...@google.com]

I can't find the discussion of why we are deprecating MediaStream.label() - can anyone point me to the right place to look? (I'm not disagreeing with the change, just curious to read)

[PhistucK]

Because it does not exist in the specification anymore (the specification link is mentioned in the intent message) -

http://w3c.github.io/mediacapture-main/getusermedia.html#idl-def-MediaStream

☆PhistucK

[Chris Harrelson]

Looks like deprecation did happen in M45 (https://codereview.chromium.org/1194763002). Please come back with an Intent to Remove before M47 with updated stats.

Thanks,

Chris

[shiv...@samsung.com]

Thanks, added new intent to remove: https://groups.google.com/a/chromium.org/forum/#!topic/blink-dev/LcZGlM9OBqs,

 also patch to remove the same is under review: https://codereview.chromium.org/1312213006/