Timed Text Working Group Teleconference

<scribe> scribe: nigel

This meeting

pal: Maybe we can close the loop on the process issue
... Also possibly the aspect ratio issue
... And review the new issue on smpte:backgroundImage in case we can close it straight away.
... So we can make some progress.

nigel: Agreed. So we'll focus on IMSC 1.
... AOB?

group: none.

IMSC issues, pull requests etc

pal: Can we start with the process question?
... In issue-92 nigel noted that the document still references the 2014 process. It was
... mentioned that we automatically adopt the new process with revisions, so it should
... be updated to 2015. I don't remember discussing it - maybe it's not an issue because
... the changes are not meaningful to us this time, but in the future, automatic adoption
... of new process documents in the middle of developing a new document could be
... really disruptive.

nigel: I think this is really important - I couldn't find the source of this assertion either.

tmichel: As I recall the only process change we approved was to 2014 but for the latest
... it was not discussed within the group.

pal: That matches my recollection too.

tmichel: Then if we want to adopt the new process we should check with the group if
... everyone is happy with it. Who requested that we use the latest process by the way?

nigel: I'm pretty sure that plh asserted that having adopted the 2014 process we by default adopt future versions as well.
... I can't see where that's written down though.

pal: I've searched through the Process document and can see no such thing.

nigel: Me too!

tmichel: It would be odd to me if it says that.

nigel: In that case it looks as though we have not agreed to adopt the 2015 process.

tmichel: Shall I investigate if we are required to move to the 2015 process or we can proceed with the 2014 one?

nigel: Yes please.

tmichel: Is the concern just for IMSC or also for TTML2 and WebVTT?

nigel: It's for everything - we need to be clear which process we're working under.
... Also by the way I'm pretty sure there are no substantive differences that would affect us with the 2015 process.

tmichel: It's not a major change like it was to adopt 2014.

nigel: There is actually a Latest Operative Version link so we need to go back and check which one we said we'd adopt: http://www.w3.org/2015/Process-20150901/

pal: Maybe it's simpler not to adopt unless we're compelled to, then we don't have to make any changes.

nigel: We have 2 choices (unless we're compelled to move to 2015):
... 1. Stay with 2014
... 2. Propose to adopt 2015, use our Decision Process and then adopt it and update our specs accordingly.

<scribe> ACTION: tmichel Investigate if we are required to move to the 2015 process [recorded in http://www.w3.org/2015/11/26-tt-minutes.html#action01]

<trackbot> Created ACTION-451 - Investigate if we are required to move to the 2015 process [on Thierry Michel - due 2015-12-03].

tmichel: I'll be able to report back on this at the next telecon

pal: You can also put the answer in issue #92, which we may be able to close directly.

nigel: We need the info in the meeting also.

tmichel: I'll first find out the situation and then we can make a decision about changing the document.

pal: Sounds good.

https://github.com/w3c/imsc/issues/92

nigel: Shall we look at the aspectRatio issue?

pal: Yes, you mentioned to me that there may be some more feedback on this.

nigel: There's nothing formal yet - a group is considering if it would be beneficial in addition to ttp:aspectRatio also to signal a safe area,
... outside of which the document contains no regions. This would allow for some scaling
... scenarios to be met that can't otherwise be handled.
... Until we receive something formal we can't formally respond though.

pal: Okay then we can probably close issue #84 https://github.com/w3c/imsc/issues/84 next week.

nigel: Have I reviewed the note for that?

pal: Yes, and I think from your comments that you're happy so we can close it next week.

nigel: Yes.

pal: The next issue is regarding smpte:backgroundImage https://github.com/w3c/imsc/issues/97

nigel: Looking at this in detail, the #image feature is defined by SMPTE-TT, but that
... has a weird referencing problem, in that it defines conformance in relation to its
... section 5.7.3, but the backgroundImage semantics are defined in a different section,
... being 5.5.3.

pal: So the bug here is in ST2052 §5.9.2 missing the reference to §5.5.2.

nigel: Agree - that's a bug in SMPTE-TT, but it doesn't help us here!

pal: We can just put a note that this bug exists.

nigel: Well that wouldn't help the formal problem.

pal: I think we're overthinking this - it's clearly enumerated as part of the vocabulary.

nigel: I'm not sure if it's a note or something more formal but I would suggest specifying
... that we mean to include the definition of smpte:backgroundImage from ST2052 §5.5.3 in the vocabulary supported by the #image extension.
... Ideally that would normative.

pal: I think it's unambiguous but if you had an issue reading it then we can add the note.

nigel: Okay I'll add a comment about this to the issue.

pal: Can we close the duplicate https://github.com/w3c/imsc/issues/93 ?

nigel: yes, no problem

<scribe> ... done.

pal: The rest are either for review or I have some actions to do for IMSC 2.

TTML and WebVTT Mapping Document

nigel: Andreas, are you going to migrate this to git?

atai: Yes, I need to check about the detail of how to do this. If I don't have enough information I'll come back to nigel and plh to ask for more.

nigel: Pierre, how did you do it for IMSC?

pal: I used a script that converts a Mercurial repo to a git repo, so I just did that.
... I created a local git repo, imported all of the IMSC 1 Mercurial repo into that, then I
... merged into that git repo the one that plh put up in github, then pushed all my changes
... back into github.
... Did plh set up the repo on github already?

nigel: Yes he did, at https://github.com/w3c/ttml-webvtt-mapping

atai: He did not merge the repo.

pal: Yes, you have to do that yourself. Create a local github repo, then use the
... Mercurial to Git script to move everything across, then separately check out the
... github repo, and merge into the local copy of the github repo the converted git repo,
... then push back upstream.

atai: That's feasible. My only question was how to restrict it.

pal: There's a git command that allows you to pick a directory and delete all others, and
... make that one a root directory. I did that after the Mercurial to git conversion.
... The command that re-roots keeps all history.

atai: Would you be able to describe this in a few lines?

pal: If you send me a reminder email I'll do that.

atai: Thank you! That's perfect.

pal: If you have branches that makes it really painful.

atai: We don't have any.
... Do we need to move anything else like the test suite?

pal: For IMSC?

nigel: They're moved already - they're in sub-directories of the top level.

AOB

atai: Pierre, Nigel and I did a presentation at IBC on TTML, EBU-TT-D and IMSC. Should we share the slides somehow?

nigel: Just natively?

atai: Maybe the easiest is to email it as a PDF to the list.

nigel: I did do that presentation as chair of TTWG, so I don't mind sharing the slides.
... Maybe the best thing is to upload to the wiki?

pal: Yes, although I'd rather avoid putting on the official TTWG wiki things that are not
... official TTWG. My recommendation is for someone to host it and send a link to the
... reflector. That would be my preference.

nigel: We did do the presentation on the EBU stand so it might make sense to see if EBU
... can host the slides also.

pal: That would be a lot cleaner.

<scribe> ACTION: nigel Check with Frans about hosting the IBC slides on the EBU space, and send a link around. [recorded in http://www.w3.org/2015/11/26-tt-minutes.html#action02]

<trackbot> Created ACTION-452 - Check with frans about hosting the ibc slides on the ebu space, and send a link around. [on Nigel Megitt - due 2015-12-03].

nigel: Looking at future meetings, I've not put any in for 24th and 31st December, with
... a restart on Jan 7. Hope that's okay with everyone.

pal: Yes, fine. I hope to publish a new CR of IMSC 1 by year's end, and see no blockers.

nigel: We need to merge all the PRs next week and agree a resolution to proceed with a
... new CR. When do the PR review periods end?

pal: Some of them end on Tuesday.

nigel: Then we can put out a call for consensus ahead of the meeting to avoid surprising
... anyone, and avoid falling foul of the Decision Process.
... Okay, that's on the agenda for next week.
... That's all for this week, thanks everyone, see you same time next week. [adjourns meeting]