MediaRecorder: update to spec (1/3)

third_party/WebKit/Source/modules/mediarecorder/MediaRecorder.idl

Index: third_party/WebKit/Source/modules/mediarecorder/MediaRecorder.idl
diff --git a/third_party/WebKit/Source/modules/mediarecorder/MediaRecorder.idl b/third_party/WebKit/Source/modules/mediarecorder/MediaRecorder.idl
index 2b92f95e497e1899a165223dba9f89df298e5c6e..a58a331f8e7016a7c5468e8ebd9079b671ee98a0 100644
--- a/third_party/WebKit/Source/modules/mediarecorder/MediaRecorder.idl
+++ b/third_party/WebKit/Source/modules/mediarecorder/MediaRecorder.idl
@@ -4,21 +4,20 @@
 
 // https://w3c.github.io/mediacapture-record/MediaRecorder.html#MediaRecorderAPI

[Peter Beverloo, 2015/12/09 23:11:29]

Mm. The GitHub repository uses both the gh-pages branch and the master branch.

Is it deliberately that they differ? AFAIK most specs only use gh-pages, since
these are just the editor's drafts anyway.

https://rawgit.com/w3c/mediacapture-record/master/MediaRecorder.html#MediaRec...
(master)
https://w3c.github.io/mediacapture-record/MediaRecorder.html#MediaRecorderAPI
(gh-pages)

[mcasas, 2015/12/10 02:16:59]

I was also uncertain about it so I pinged the appropriate 
savvy fellas in the w3c and they said both approaches
are reasonable, WebRtc and GetUserMedia specs being
maintained that way master->gh-pages->w3c...

Personally just wanted to have a way of bundling
a few minor corrections before publishing to 
gh-pages and then to w3c. I'm too used to having
several channels I guess :) Moving on, I guess I'll
it's reasonable to have both much closer.

 
-enum RecordingStateEnum { "inactive", "recording", "paused" };
+enum RecordingState { "inactive", "recording", "paused" };
 
 [
     GarbageCollected,
     ActiveDOMObject,
     // TODO(mcasas): consider changing |mimeType| to a dictionary with said key, see https://github.com/w3c/mediacapture-record/issues/19

[Peter Beverloo, 2015/12/09 23:11:29]

I see that Philip also pointed this out, but since Gecko already does this per
the issue it's probably fine to go ahead and match their implementation.

Doesn't have to be in this CL of course :).

[mcasas, 2015/12/10 02:16:59]

I agree and is being done here: http://crrev.com/1507183002/ ;)
(which will come your way soon)

-    // TODO(mcasas): Remove [TreatUndefinedAs], see https://github.com/w3c/mediacapture-record/issues/7
-    Constructor(MediaStream stream, [TreatUndefinedAs=Missing] optional DOMString mimeType),
+    Constructor(MediaStream stream, optional DOMString mimeType),
     ConstructorCallWith=ExecutionContext,
     RaisesException=Constructor,
     RuntimeEnabled=MediaRecorder,
 ] interface MediaRecorder : EventTarget {
     readonly attribute MediaStream stream;
     readonly attribute DOMString mimeType;
-    readonly attribute RecordingStateEnum state;
+    readonly attribute RecordingState state;
     attribute boolean ignoreMutedMedia;

[Peter Beverloo, 2015/12/09 23:11:29]

micro nit: it's easier to confirm how well we match the specification's IDL if
we match the declaration order. You could consider moving this attribute to come
after the events.

[mcasas, 2015/12/10 02:16:59]

Done.

 
     attribute EventHandler onstart;