Timed Text Working Group Teleconference

02 Mar 2017

See also: IRC log


Dae, Glenn, Nigel, Pierre, Thierry, Andreas, Mike


<scribe> Scribe: Nigel

This Meeting

Nigel: Today, we have a proposal on meeting times, one IMSC issue and some TTML issues.
... Any other topics to cover?

Andreas: I'd like to cover the pull request on TTML1.

Meetings times

Dae: I think there's enough to talk about between now and the end of the month I propose
... to start "early", at 10am Boston time, so we can have longer meetings.

Nigel: By the way, DST comes in, in the US, soon, and at the end of March in Europe.

Thierry: I believe it is on the 12th in the US and the 26th in Europe.
... So there is an impact in Europe, but we are back in sync after March 12th.

Nigel: Dae, I think this is a good suggestion.

Andreas: It's fine for me to start an hour earlier but I cannot allocate 2 hours to the meeting, however this is not a problem if you go ahead.

Nigel: Okay as Chair I'm going to make the call and say we will do this, begin at 10am Boston and set a duration of 2 hours for all the remaining meetings in March.


foreign namespace usage is underspecified

Mike: Regarding the example of smpte:information my view is that it is an error in the
... SMPTE-TT spec, and it should say "descendant" of the head element, with the intent that
... it be made a child of the `metadata` element under `head`.

Glenn: The reason we put `metadata` everywhere was to support the addition of foreign
... namespace data, with the exception of the `tt` element on the basis that such metadata
... could be placed under `head/metadata`.
... It is a bit indirect but if you follow through TTML1 from §3 to §4 to §5 the foreign
... namespace elements _are_ pruned because their names are not present in the abstract
... document type.

Nigel: That's an argument contrary to my reading, but I can accept that.

Andreas: That matches my understanding too.

group: [seems to have consensus that foreign namespace elements are prohibited in TTML1 except as children of metadata]

Nigel: My concern for the concrete encoding in IMSC1 is that we may prohibit the addition
... of vocabulary that is orthogonal to processing defined in TTML and IMSC and that may
... prevent some useful use cases.

Mike: It seems like an arbitrary syntactical restriction since foreign namespace attributes
... are allowed - why allow attributes but not elements?

Glenn: That was a deliberate choice to avoid putting TTML namespace elements as children
... of foreign namespace elements.
... A clarification may be useful in TTML1 and possibly a normative statement in TTML2 would be advisable.

Nigel: Is everyone except me here comfortable with prohibiting foreign namespace elements in IMSC1 but permitting foreign namespace attributes?

Mike: I don't want to add further restrictions but I think that is what TTML1 says today.
... I want to be crystal clear in IMSC1, and then we can take up the question in TTML1 and TTML2.

group: Consensus to clarify in IMSC 1 that TTML1 permits foreign namespace elements only in tt:metadata and attributes everywhere, §6.2 of IMSC1.

Pierre: There's already a note referring to structural elements of TTML1, so it could be an
... extension to that. We may want to go further, or stay silent, I'm not sure which yet.

Nigel: Who will take this action?

Pierre: I will do this, and check in with Mike before posting as a PR.

Mike: I'm relieved of the action from last week.


Dae: If anyone is concerned that we cannot meet the TTML2 WR deadline please raise any
... issues that will block us from getting there.

Pierre: Movielabs' position is that all issues have to be resolved before by WR, either by closing or deferring them.
... So a priori it means every issue that has not been assigned by consensus to TTML.next.

Dae: For example the issues raised by i18n review.

Pierre: I think anyone in the group can take a stab at resolving those issues.
... There are some that Movielabs has stronger opinions on but in general it is not possible
... to defer them en masse.

Nigel: The horizontal review comments should not be ignored but we may be able to address
... them after WR and before CR.

Pierre: I agree with that.

Nigel: I need to run now but the point is made Dae that we have issues to cover.

Dae: We should focus on the ones that are not 'easy' first.

Glenn: I don't intend to address the HR comments before WR by the way, so we need to
... discuss what to do about those if there is not consensus on that.

Nigel: Thanks everyone. [adjourns meeting]

Summary of Action Items

Summary of Resolutions

[End of minutes]

Minutes formatted by David Booth's scribe.perl version 1.152 (CVS log)
$Date: 2017/03/02 18:40:30 $