See also: IRC log
<cyril> scribe: cyril
nigel: no objection or addition
to the agenda?
... agenda is approved as posted
gkatsev: I started looking at the preliminary impl report and I can give an overview
nigel: there are 2 PR and one issue marked for agenda
<glenn> https://github.com/w3c/tt-profile-registry/pull/55
github: https://github.com/w3c/tt-profile-registry/pull/55
nigel: there is one particular
comment that need discussion
... It's about the IANA registration part
... this PR adds text that says the document augments the
registration
... but this document defines the MIME type and cannot augment
it
glenn: I looked at it a bit more
and now agree with you
... we can change without IANA approval afaik
nigel: afaik
glenn: then I will edit it and should be able to resolve that
nigel: is it worth going over any other part of this PR
glenn: no
nigel: then we'll continue the discussion about how we reference other specs offline
summary: editor has a way forward to deal with review comments on IANA registration
github: https://github.com/w3c/tt-profile-registry/pull/57
nigel: this PR creates a JSON in
the repository to make it easier to view PR
... it's enabled on all repositories of W3C but it needs this
file
glenn: I'll approve that, I don't know what previewing a PR means
nigel: at the top of the PR, preview and diff links get added
glenn: I don't think this is enable in TTML2
nigel: it works for respec and bikeshed specs, not for more complicated specs
SUMMARY: Glenn will approve and merge it
github-bot: https://github.com/w3c/tt-profile-registry/issues/56
<github-bot> cyril, Sorry, I don't understand that command. Try 'help'.
github-bot: https://github.com/w3c/tt-profile-registry/issue/56
<github-bot> cyril, Sorry, I don't understand that command. Try 'help'.
github: https://github.com/w3c/tt-profile-registry/issues/56
atai2: we need to double-check the reference
nigel: maybe the etx2 is wrong
atai2: it points to 3350
... it's not wrong, designator and identifier is correct
... what may need to be adjusted is the link to the
specification
... but I need to double check it
nigel: in the current editor's
draft, there are version links to the 1-0 and 1-2 but there is
no link to 1-1.
... it could be my mistake or something is wrong in the EBU
specs
atai2: I think you are
right
... there should be EBU-TT 1.1
SUMMARY: Nigel will open a PR to modify ext2 to point to EBU-TT 1.1
nigel: we have nothing labelled
for agenda in the issues
... for the things I've raised I owe the group a bit more
details
... I'll provide more until the next meeting
... no one seems to want to add more now, action everybody to
continue reviewing the current issues and ask for details
<nigel> Meeting wiki page
nigel: according to the wiki, we
have 4 people listed as attending
... thank you andreas for the doodle for the dinner
... thank you thierry for updating the wiki page
... any admin questions?
cyril: I won't able to
travel
... will there be a way to join in
nigel: no problem in setting up a
webex
... Frans already set up of webex for the friday morning
... I will set up one for the thursday meeting
... I've given myself action 16
... also I want to mention that the joint meeting with the EBU
timed text on the friday morning will discuss live
contributions
... there is a proposal that 2 people will report on their
implementation experience: Matt Simpson and me
... we shouldn't spend more than 30 min on these 2 slots
... do people have things to contribute?
... If you have any idea, please get in touch with me
... the other slightly admin point is that we'll do it as a W3C
meeting in terms of IPR
glenn: on impl experience, are you limiting it to the live scope?
nigel: it is specifically on live contributions
atai2: speaking as a EBU
co-chair, we also want to discuss how the EBU TT group and the
W3C TT group can collaborate
... we have most of the members in one room
... we should discuss what should be done in w3c and what
should be done in ebu tt
... what should moved to w3c if any
... I want to make it one topic
cyril: I would also be interested
in discussing what can be done for the EBU features that are in
IMSC and not in TTML2
... the fact that IMSC1.1 is not a strict subset of TTML2
... for example, discussing if copyright of EBU TT features
could be transfered to W3C
atai2: it would be good to discuss if the EBU TT group thinks extension will be done in the future in EBU or should be done in W3C
nigel: a slighlty broader point
is that TTML is extensible but what happens when an extension
needs to be adopted more broadly
... I would like to understand why any change is needed
... it may not be the most elegant thing (several namespaces)
but it's not that bad
... we need to motivate a change
cyril: there are 2 parts: one is the fact that we have multiple namespaces but the first point is having one place to find the specification
nigel: we also want to modularize
cyril: it's harder to implement a spec if you have to pull sub-specs from different orgs
nigel: I can go ahead and edit
the wiki for the agenda
... anything else?
... no
nigel: I don't think there is anything to discuss
glenn: just a clarification, you
mentioned future editorial changes
... I'm distinguishing between new features and substantive
changes to existing features
... the issues I've been filing recently come from my internal
list made during the finalization of the 1st edition
... I have about a dozen more
... hope to finish that before the meeting if we want to
discuss some of them
nigel: they are indeed editing tasks, not necessarily editorial
pal: on that TTML2 topic, we had
discussed requiring test cases to be submitted with
tickets
... we should stick to that
glenn: I don't recall that
... this is not a bad idea
nigel: that would apply to the
substantive changes
... a test case would be very helpful
pal: it helps folks testing their implementation quickly
glenn: also important for when we need to move to PR
nigel: it also helps for the review of the spec
glenn: makes it more concrete
RESOLUTION: we reaffirm that we want to have test vectors for substantive changes
nigel: gkatsev wants to give an update
gkatsev: I've started looking a
bit more
... big initiative on wpt.fyi
... has an interop section
... shows you which test passes in 1, 2 or 3 browsers
... for the things that can be tested automatically, I'll use
that
<nigel> web-platform-tests dashboard for webvtt
gkatsev: overall 80% of the
features are implemented in 2 or more browsers
... there are some features for which I'm not sure
... for those that won't pass, I'll discuss with Silvia
... but 80% seems not as bas as I thought
cyril: what about the new features: regions, styles
gkatsev: safari 12.1 is supposed
to support regions
... VLC supports it
... those 2 should qualify as implementation
nigel: can you clarify how you reached the number of 80%
gkatsev: I ignored rendering
because they do not run automatically
... the value reported on wpt.fyi is wrong
atai
cyril
atai2: do you have an idea if the WPT fully covers the version of WebVTT
gkatsev: I am not 100% sure but
it seems the coverage is very good
... but we need to make sure it is before we can rely on
WPT
atai2: I'm not sure when the
tests in WPT were made
... we should check if changes to the spec were made after the
tests were added to WPT
tmichel
tmichel: I am wondering about the
coloring of the cells
... light green vs yellow
gkatsev: I assumed the darker the
green the higher the % of passing tests
... I think region is the most recent and has more risk
tmichel: you mentioned VLC as a
pretty good candidate for region support
... how do we test that?
gkatsev: we can't test it
automatically but we can load a VTT file in VLC and see how it
work
... we can't test it automatically but we can load a VTT file
in VLC and see how it works
<nigel> scribe: nigel