See also: IRC log
<scribe> scribe: nigel
Glenn: I have to leave at 40 past the hour.
Nigel: For today I think we want
to focus on TTML, then HDR in PNG and IMSC, plus TPAC
... Any other business? Or any specific points to raise?
group: [no other business]
Nigel: Thank you everyone for
contributing to the extra volume of activity in the past
week.
... Progress tracking
... Some people have taken on issues, some pull requests have
been merged, and proposed.
... Does anyone want to assign themselves to an issue but
discuss it beforehand?
group: [none]
Glenn: I think myself and Pierre
have self-assigned enough that we don't need any more
... right now. This month I'm focusing on getting everything
out the door so I'm on a steam
... roller and intend to stay on it right through the
month.
... There have been a couple of comments about labelling HR
comments as WR. Last night
... I pointed out that as long as there are other open HR
issues on the WR list we can proceed
... with handling those comments, however at some point we may
have the current HR
... issues open. If there are other issues open we may have to
discuss holding off the WR
... but we don't know that yet.
Pierre: We've had no discussion on labelling issues.
Glenn: Right now I'm using the WR
milestone so I'd like things not to be added without
... discussing them first unless you have a PR, in which case
feel free to put it on WR.
Nigel: Can we back up a level - are we talking about a milestone or a label?
Glenn: We're talking about a
milestone - it detracts from my own processing
particularly
... if there's no pull request on it. If issues have a pull
request they can add it to the milestone.
Pierre: You can just filter
issues assigned to you.
... A lot of issues are labelled as WR without being assigned,
so I don't follow you.
... We've agreed to take a group effort so you can just assign
issues to yourself and use that
... to filter on.
Glenn: you can create a new
milestone TTML2WR-HR and when there's a PR move it across
... to the TTML2WR milestone.
Pierre: Milestones are for deciding which issues need to be resolved before a release.
Nigel: Stepping aside from the
assignee filtering it is true that we may in principle
not
... address all the HR issues in the WR, though accepting that
Movielabs does want to
... resolve them for WR. So I believe that the TTML2WR
milestone is correct (and it coincides
... with my understanding of the group's agreement, though I
can't recall the minutes for
... that) and it is reasonable to move things into TTML2WR
milestone if we know we have
... a resolution for them.
David_Ronca: I'm going to have to
push back on Glenn - we're not happy with the progress
... on TTML2 so we need to have confidence that we are going to
finish. Having a milestone
... used as somebody's personal review tool isn't appropriate.
I believe that all the issues
... that we believe must be resolved into WR should be tracked
through a milestone. That's
... the way I would do it in a software project.
Glenn: Right now the Editor is serving as the project manager for this particular document.
Nigel: If anyone thinks an issue
is mis-attributed to a milestone then please raise that
for
... discussion.
Glenn: My point was that I'm
effectively serving as the programme manager for TTML2 by
... default as the Editor. If the group wants to review every
issue and decide on what's on
... that milestone list then we could do that but I think that
would not be a good use of time.
David_Ronca: I just think that
the milestone list should just be those issues that need
to
... be resolved by WR.
Nigel: All the issues that are
not HR comments are in the TTML2WR milestone. So we have
... only two cases to cover:
... 1. Issues that we in fact do not want to fix in
TTML2WR.
... 2. HR/WR issues that have no resolution proposed but need
to be fixed in TTML2WR.
... My understanding was this was the position as agreed in
London.
... This is not excluding from resolution any issues not
currently in the TTML2WR milestone.
... In that case just resolve the issue and then we can move
the issue to TTML2WR milestone.
Pierre: I don't agree with this
and ask to put it on the agenda for next week so the
group
... can form a consensus opinion.
Nigel: Okay I will table that for
next week.
... Okay let's stop this discussion here please.
... Now what pull requests or issues should we discuss?
Glenn: There are 2 pull requests currently open.
Updated tts:writingMode example (issue #269)
Glenn: I'm wondering if we should
change the existing example (and in TTML1) or add a new
... example. There are two problems with the example. Firstly
showing the upright
... orientation of roman text, which in TTML2 would be set
sideways. To achieve that example
... output (in the current example) we would have to add a
textOrientation="upright"
... property. In TTML1 we did not have that property. I would
speculate that there are few
... if any implementations that do what is speculated in that
original example.
... The 2nd part of the issue is the bidi part. When WGBH
prepared the current example and
... image in TTML1 for the bidi portion there weren't any bidi
implementations around so I
... think Geoff did something simple which was to reverse the
horizontal lettering. I
... support updating that, which would be better for TTML1 and
TTML2. But we do have this
... technical issue on the vertical portion.
Nigel: Please can you add a note
to the issue/pull request saying exactly what you think
... needs tweaking?
Glenn: Yes I will do that. Thanks Pierre for moving some of this forward, I appreciate it.
Adds xml:lang to metadata examples (issue #271)
Nigel: I made an alternative outline proposal on this which Richard gave a +1 to. Pierre, would that work for you?
Pierre: Yes, I'll reassign it to you if that's ok?
Nigel: Ok, yes.
Pierre: [assigns it]
Nigel: Thank you.
Pierre: I think we need to revert
a commit for the bopomofo - there was an approved PR
... which was modified before merge and then there were
comments.
Glenn: I took the liberty to change it before merging it.
Pierre: But you approved the change, and then merged it before the 3 days?
Glenn: We don't want to start the clock again after every change.
Nigel: Actually we do, however this is editorial so the process allows for this.
Pierre: This seems unproductive and ineffiicient so I can't imagine why we want to do that [modify an approved pull request before merging].
Nigel: Can I just check if the note as currently committed is okay for you?
Pierre: I'm not an expert but the
term that was removed is used everywhere so it seems
... safest to keep it, especially as the term was used by the
original commenter. It's really
... the process that I'm worried about here because it's going
to affect us in the future.
Glenn: The term zhuyin fuhao is
technically accurate but not generally used except by the
... Chinese, and is not used in CSS or Unicode.
Pierre: Typically not resolving
comments doesn't make them go away but costs more in the
... future.
Nigel: In this particular case we
can simply check in with the original commenter, Richard,
... and see if he is happy with the resolution. I don't think
we need to worry about hypothetical
... future review comments, only ones that we have actually
received.
... I've added a comment to the original issue (#253)
requesting feedback from Richard,
... and highlighting the discussion we have had, and reopened
the issue.
... Are there any other issues that anyone wants to raise for
discussion even in the absence
... of Glenn (who has left)?
David_Ronca: We will start looking at the Ruby and Japanese HR issues
David_Ronca: [drops off the call with thanks]
Nigel: The IMSC 1.0.1 review
period will end on Sunday. The next thing will be the
tests
... for CR exit criteria.
Pierre: I haven't started on that
really; my next step is to see if Andreas and I can
create
... more examples for lineGap.
Nigel: We have had some WR
comments in, one that I'll just forward now with
apologies
... for the delay:
https://lists.w3.org/Archives/Member/member-tt/2017May/0000.html
Pierre: I'd like to enter those
comments on GitHub but they were sent privately, so how
do
... we do that? Can you ask them if they are happy for the
comments to be made public
... otherwise they will be harder to track. Or we can enter
them without attribution.
Nigel: Okay I will do that.
... By the way we had a second response:
https://lists.w3.org/Archives/Member/member-tt/2017Apr/0004.html
Nigel: My proposal would be wait
until after the deadline has passed before discussing
... each comment.
Pierre: I just found a bug and fixed it in the test vector, I don't think there's anything else.
https://github.com/w3c/imsc-tests
Nigel: Do they need to be reviewed?
Pierre: I'm just going to go with
the 14 days and if there are no comments then I'll merge
... them. We are not under time pressure here.
Nigel: Thanks, I'll take a look at those.
Pierre: We need to set a date for closing review on this.
Nigel: It's worth mentioning that
Thierry asked Chris Lilley if he could come back with any
... comments via the Color CG in 2 weeks. Chris hasn't
responded yet.
Pierre: I'm happy to just say 4
weeks, fix something and then go forward with it until
... someone complains.
Nigel: I don't want this to get in the way of other work that we're doing.
Pierre: I'm also happy with Thierry's proposal to publish now and then update.
Nigel: I think we should publish
in 3 weeks, so a Resolution to publish next week, then
... 2 more weeks. That should give plenty of time for Chris to
get back to us if there's
... anything needed before first publication. Otherwise we can
modify the document later
... with little process requirement.
Nigel: [reviews the WBS responses
as in the agenda]
... Okay let's close for today. Thanks all.
Mike: Regrets from me for the next 2 weeks.
Nigel: Okay, thanks Mike. [adjourns meeting]