- From: Timed Text Working Group Issue Tracker <sysbot+tracker@w3.org>
- Date: Wed, 15 Feb 2012 12:39:27 +0000
- To: public-tt@w3.org
ISSUE-154 (endsync behavior): Clarify the endsync behavior [DFXP 1.0] http://www.w3.org/AudioVideo/TT/tracker/issues/154 Raised by: Sean Hayes On product: DFXP 1.0 Summary: Time containers should have a default value of endsync=’all’ rather than the current default endsync=’last’, otherwise marker mode does not work. Detail: In SMPTE discontinuous mode TTML defines time containers to use the marker mode (event based) timing model of SMIL; however in this mode the elements inside a time container are not waiting for events unless the parent timecontainer is active. However unless the parent duration is explicitly notated, since all of the children are event based they would under SMIL have no resolved active time, which in the default endsync model results in a parent with zero duration. I think we can most simply resolve this issue by defining that time containers have a default value of endsync=’all’ rather than the SMIL default endsync=’last’. This means that the parent will stay around until all of the children have completed at least one activation, which is a more intuitive expectation for a caption file anyway.
Received on Wednesday, 15 February 2012 12:39:29 UTC