Right now WebKit uses a custom video sink which only task is to pass, through a GObject singal, the decoded video frame to WebKit. Replacing this custom video sink with a GstAppSink, which -at first sight- does exactly the same things, we could reduce the code to maintain in WebKit.
Created attachment 192063 [details] WIP: patch for replace the custom WebKitVideoSink for AppSink I'm worried about the A/V sync with large buffers
Created attachment 192904 [details] Patch
Attachment 192904 [details] did not pass style-queue: Failed to run "['Tools/Scripts/check-webkit-style', '--diff-files', u'Source/WebCore/ChangeLog', u'Source/WebCore/GNUmakefile.list.am', u'Source/WebCore/PlatformEfl.cmake', u'Source/WebCore/Target.pri', u'Source/WebCore/WebCore.gypi', u'Source/WebCore/platform/graphics/gstreamer/GRefPtrGStreamer.cpp', u'Source/WebCore/platform/graphics/gstreamer/GRefPtrGStreamer.h', u'Source/WebCore/platform/graphics/gstreamer/MediaPlayerPrivateGStreamerBase.cpp', u'Source/WebCore/platform/graphics/gstreamer/MediaPlayerPrivateGStreamerBase.h', u'Source/WebCore/platform/graphics/gstreamer/VideoSinkGStreamer.cpp', u'Source/WebCore/platform/graphics/gstreamer/VideoSinkGStreamer.h', u'Tools/ChangeLog', u'Tools/Scripts/webkitpy/style/checker.py', u'Tools/Scripts/webkitpy/style/checker_unittest.py']" exit_code: 1 Source/WebCore/platform/graphics/gstreamer/MediaPlayerPrivateGStreamerBase.cpp:448: Tests for true/false, null/non-null, and zero/non-zero should all be done without equality comparisons. [readability/comparison_to_zero] [5] Total errors found: 1 in 6 files If any of these errors are false positives, please file a bug against check-webkit-style.
Sebastian, can you please provide some feedback about this patch?
By using appsink you won't be able to handle the GstVideoMeta/GstVideoCropMeta, and will also be unable to handle special memory (think: EGLImage, VAAPI, ...) properly.
(In reply to comment #5) > By using appsink you won't be able to handle the GstVideoMeta/GstVideoCropMeta, and will also be unable to handle special memory (think: EGLImage, VAAPI, ...) properly. Doh! Thanks! Phil, do I change the status to "resolved" or something?
Yeah, let's close this then.