The fitness score calculated for constraints passed to getUserMedia is not used when selecting between multiple devices.
<rdar://problem/31138936>
Created attachment 304910 [details] Proposed patch.
Comment on attachment 304910 [details] Proposed patch. View in context: https://bugs.webkit.org/attachment.cgi?id=304910&action=review > LayoutTests/fast/mediastream/MediaDevices-enumerateDevices-expected.txt:31 > +FAIL deviceIds.indexOf(captureDevice.deviceId) should not be -1. Is this an expected failure? > LayoutTests/fast/mediastream/MediaDevices-enumerateDevices-expected.txt:38 > +FAIL deviceIds.indexOf(captureDevice.deviceId) should not be -1. Ditto > LayoutTests/fast/mediastream/MediaDevices-enumerateDevices-expected.txt:64 > +FAIL deviceIds.indexOf(captureDevice.deviceId) should not be -1. Ditto > LayoutTests/fast/mediastream/MediaDevices-enumerateDevices-expected.txt:71 > +FAIL deviceIds.indexOf(captureDevice.deviceId) should not be -1. Ditto > LayoutTests/fast/mediastream/MediaDevices-enumerateDevices-expected.txt:97 > +FAIL deviceIds.indexOf(captureDevice.deviceId) should not be -1. Ditto > LayoutTests/fast/mediastream/MediaDevices-enumerateDevices-expected.txt:104 > +FAIL deviceIds.indexOf(captureDevice.deviceId) should not be -1. Ditto. If so, we should come up with a way of reporting it as a "PASS, failed expectedly."
Comment on attachment 304910 [details] Proposed patch. View in context: https://bugs.webkit.org/attachment.cgi?id=304910&action=review >> LayoutTests/fast/mediastream/MediaDevices-enumerateDevices-expected.txt:104 >> +FAIL deviceIds.indexOf(captureDevice.deviceId) should not be -1. > > Ditto. > > If so, we should come up with a way of reporting it as a "PASS, failed expectedly." Oops, I need to update the test and results now that there are four mock devices. Good catch!
Created attachment 304929 [details] Patch for landing.
Comment on attachment 304929 [details] Patch for landing. Clearing flags on attachment: 304929 Committed r214178: <http://trac.webkit.org/changeset/214178>