Let's discuss the web speech synthesis for ports using gstreamer here. We have a working branch: https://github.com/shivamidow/WebKit/tree/eng/speech-synthesis-with-gst-flite This patch passed 22 / 26 webkit tests in fast/speechsynthesis. 4 test failed due to crash and unknown reasons. We need to investigate it more.
I don't want to delay this patch for further bug fixes. It is itself a huge chunk so I suggest splitting the web speech synthesis support into some phases. Phase 1. Support build and basic functionalities (e.g., play, stop, pause, cancel) -> the current patch Phase 2. Fix bugs on layout tests (fast/speechsynthesis) and enable the tests Phase 3. Further bug fixes as we find new ones. @philn, thought?
Or we can enable fast/speechsynthesis in phase 1 but add failed 4 tests as exceptions to fix later.
> Phase 1. Support build and basic functionalities (e.g., play, stop, pause, cancel) -> the current patch Without GStreamer usage from UIProcess?
(In reply to Philippe Normand from comment #3) > > Phase 1. Support build and basic functionalities (e.g., play, stop, pause, cancel) -> the current patch > > Without GStreamer usage from UIProcess? Yes. That was fixed already. =)
Pull request: https://github.com/WebKit/WebKit/pull/8677
Committed 259197@main (f99efce2948a): <https://commits.webkit.org/259197@main> Reviewed commits have been landed. Closing PR #8677 and removing active labels.