Modifications on GstMatroskaDemux tracked in GStreamer bugzilla: https://bugzilla.gnome.org/show_bug.cgi?id=765275 https://bugzilla.gnome.org/attachment.cgi?id=365211
Created attachment 348764 [details] Patch
Looks good, I'll let Charlie have a look as well and we can land this here when it does upstream or it is needed for something else.
Created attachment 351538 [details] Patch
Comment on attachment 351538 [details] Patch We need changes of bug 190253 as well.
Created attachment 351670 [details] Patch
Comment on attachment 351670 [details] Patch Clearing flags on attachment: 351670 Committed r236911: <https://trac.webkit.org/changeset/236911>
All reviewed patches have been landed. Closing bug.
<rdar://problem/45080591>
Why are those patches needed? I'd rather flag related layout tests than adding them, to be honest. The use of GST_PROTECTION_UNSPECIFIED_SYSTEM_ID without wrapping ifdef broke the build BTW.
https://bugs.webkit.org/show_bug.cgi?id=190471
(In reply to Philippe Normand from comment #9) > Why are those patches needed? I'd rather flag related layout tests than > adding them, to be honest. They are needed for encrypted WebM. media/encrypted-media/clearKey/clearKey-encrypted-webm-event-mse.html media/encrypted-media/clearKey/clearKey-webm-video-playback-mse.html > > The use of GST_PROTECTION_UNSPECIFIED_SYSTEM_ID without wrapping ifdef broke > the build BTW. I don't understand how it broke the build, if you have updated your jhbuild.
(In reply to Yacine Bandou from comment #11) > (In reply to Philippe Normand from comment #9) > > Why are those patches needed? I'd rather flag related layout tests than > > adding them, to be honest. > > They are needed for encrypted WebM. > > media/encrypted-media/clearKey/clearKey-encrypted-webm-event-mse.html > media/encrypted-media/clearKey/clearKey-webm-video-playback-mse.html > Can't we flag/skip these two until gst 1.16 is released? By the end of the year hopefully. > > > > The use of GST_PROTECTION_UNSPECIFIED_SYSTEM_ID without wrapping ifdef broke > > the build BTW. > > I don't understand how it broke the build, if you have updated your jhbuild. Because I didn't use jhbuild. Whenever possible we should try to avoid frankensteining our jhbuild with patches, especially those adding new API.
(In reply to Philippe Normand from comment #12) > (In reply to Yacine Bandou from comment #11) > > (In reply to Philippe Normand from comment #9) > > > > I don't understand how it broke the build, if you have updated your jhbuild. > > Because I didn't use jhbuild. Whenever possible we should try to avoid > frankensteining our jhbuild with patches, especially those adding new API. Understood.