Closed spoeschel closed 7 years ago
Shouldn't the two EBU-TT fields inherit the corresponding values from the EBU STL file instead of being set/overwritten with the current date? As long as the content of the original file isn't altered this seems to make more sens to me and it would be essential for round tripping reasons. Am I missing something?
The respective fields from EBU STL are already covered by #34; they are stored in separate metadata fields.
THX!
Just for completeness: ebuttm:documentCreationDate
and ebuttm:documentRevisionDate
in EBU-TT Part 1 version 1.0 were indeed intended to be the corresponding fields for STL Creation Date and STL Revision Number. But there was the conflict that the fields in the description were meant to be the creation data and revision number of the EBU-TT XML document, so the STL metatadata would be overwritten.
The workaround that was found in EBU-TT Part 2 version 0.9 to map the STL GSI fields to the ebuttExt:
namespace. As currently SCF is based on EBU-TT Part 1 version 1.0 and EBU-TT Part 2 version 0.9 we follow this strategy.
When SCF will be based on the latest specs (so a later version of EBU-TT Part 1 and the 1.0 version of EBU-TT Part 2) this will again be changed. EBU-TT Part 1 version 1.1 defines ebuttm:stlCreationDate
and ebuttm:stlRevisionNumber
and SCF will use these new fields. The STL Mapping guide (EBU-TT Part 2) is not published yet as 1.0 but is expected to be published this year.
The two EBU-TT fields
ebuttm:documentCreationDate
andebuttm:documentRevisionDate
shall be set/initialized with the current date.