{minutes} TTWG Meeting 2016-05-05

Previous Topic Next Topic
classic Classic list List threaded Threaded
1 message Options
Reply | Threaded
Open this post in threaded view

{minutes} TTWG Meeting 2016-05-05

Thanks all for a very productive extended meeting today. Minutes can be found in HTML format at https://www.w3.org/2016/05/05-tt-minutes.html

We made 2 resolutions today:

RESOLUTION: Publish http://rawgit.com/w3c/ttml2/ttml2-WD-201605XX/spec/ttml2.html as a new WD of TTML2

RESOLUTION: We will publish the TTML Profiles Registry as a WG Note on /TR and enable automated publication of updates following PR review and consensus.

The review period for each of these resolutions, according to our Decision Policy, expires on 19th May 2016.

Minutes in text format:


      [1] http://www.w3.org/

                Timed Text Working Group Teleconference

05 May 2016

   See also: [2]IRC log

      [2] http://www.w3.org/2016/05/05-tt-irc


          Coralie, plh, Glenn, Nigel, Mike, Pierre, Andreas

          Thierry, Frans




     * [3]Topics
         1. [4]This Meeting
         2. [5]IMSC
         3. [6]TTML2
         4. [7]Profiles registry
         5. [8]TPAC
     * [9]Summary of Action Items
     * [10]Summary of Resolutions

   <scribe> scribe: nigel

This Meeting

   nigel: Runs through agenda - will cover IMSC PR first, then
   also TPAC, TTML2 (WD publication) and Profiles registry issues
   ... AOB?

   Andreas: I could give an update on the mapping document if time

   nigel: The Webex is the same as the previous one if anyone
   wants to add it to their diaries.
   ... It has just been extended so the recurrence no longer ends.



   <trackbot> action-465 -- Thierry Michel to Ping plh privately
   regarding imsc rec press release -- due 2016-05-05 -- OPEN


     [11] http://www.w3.org/AudioVideo/TT/tracker/actions/465

   close action-465

   <trackbot> Closed action-465.

   nigel: There's been some discomfort around the press release
   for IMSC. Coralie, can I hand over to you?

   koalie: Hi I'm Coralie Mercier. A number of people have been in
   touch with me and Karen.
   ... Unfortunately we were not able to publish the PR that was
   planned. Some other things
   ... that occurred resulted in this being deprioritised. Also I
   was away in China behind the
   ... firewall and unaware of what was happening.
   ... Karen and I met: the materials we gathered for the Press
   Release were substantial.
   ... It was a great opportunity to publish on IMSC 1. I
   apologise that we could not do so.
   ... We still believe we have a story to tell but the timing is
   not so great.
   ... What I want to take away is what are our next good
   opportunities for this?

   mike: Thanks for the background and history. The concern from
   my perspective is not so
   ... much whether the PR was important to W3C but rather that
   there was a proactive
   ... reach out to 3rd parties to provide material, from
   organisations that were not all part
   ... of W3C. To learn that the PR may never be issued is the
   primary concern. Had there
   ... not been the reach out to 3rd parties it would not be so
   important. W3C should
   ... issue this at some point in the future to save face,
   otherwise others will not provide
   ... material in the future.

   koalie: Thank you Mike, I think we're in agreement.

   pal: In that case the next step is to figure out schedule.

   nigel: Do we have any other opportunities, maybe not standalone
   press release?

   koalie: I don't know. The best case is a natural trigger to do
   that press release. This is
   ... why I'm on the call. Is there something planned in the near
   future? Otherwise Karen and I need to figure out which
   secondary trigger we could use

   nigel: I don't think there is an immediate future trigger.

   glenn: We had the trigger and didn't make the date. Now we just
   need to get the release out.

   mike: We may also recommend that TTML2 go to a new public WD.
   Possibly that could
   ... be a trigger?
   ... If not, what Glenn said.

   koalie: That would be the sort of trigger I had in mind.

   nigel: I'm concerned that publishing info about IMSC 1
   associated with a TTML2 WD could be confusing.

   glenn: If you issue a PR primarily re IMSC 1 and then also add
   a note to the end about TTML2 WD that would serve the purpose.
   ... And Pierre could add a blurb about the plans for IMSC
   relating to TTML2

   pal: So the idea is primarily IMSC 1 with a note about TTML2
   and plans for IMSC relating to TTML2?

   nigel: Yes

   <koalie> [sounds fine to me as well, thanks]

   pal: Sounds fine to me.

   nigel: I think the next steps then are for some new words to be
   added for TTML2 and
   ... IMSC, from Glenn and Pierre. Is the rest of the PR text

   koalie: Yes, it's ready.

   pal: If you send me the text then I'll make the changes.

   <koalie> [Coralie leaves]

   glenn: I'll provide some words also.


   <trackbot> action-466 -- Nigel Megitt to Review notes to see
   who if anyone should be added to imsc acknowledgements for
   special thanks. -- due 2016-05-05 -- OPEN


     [12] http://www.w3.org/AudioVideo/TT/tracker/actions/466

   nigel: I've not been able to make progress on this yet.


     [13] https://github.com/w3c/imsc/pull/180

   glenn: I didn't see the need to attribute organisations, but
   would accept them.

   nigel: I think we have consensus both ways; as chair I have a
   preference for including affiliations so I'd propose we do
   ... The current draft doesn't list WG members but does list
   some specific contributors
   ... and affiliations, with "Invited Experts" for non-member
   affiliations, for John and Mike.
   ... Mike, is that okay with you?

   mike: Yes, that's reasonable.

   pal: [checks the wording on the DECE]

   mike: [happy with the wording]

   pal: The patch looks good to me.

   glenn: So you don't want to simplify it further according to my

   nigel: My preference is to keep the affiliations, since we
   should also recognise the
   ... organisations that provide resources.

   pal: The path of least resistance is to merge the PR.

   glenn: Can I ask for Skynav Inc to be changed to Skynav?

   atai: I'm happy with IRT or the full name to match "British
   Broadcasting Corporation"

   pal: So we'll change IRT to full name and remove Inc from
   ... Okay I'll do that right now.

   nigel: Institut für Rundfunktechnik

   pal: I've committed to the branch
   ... Shall we accept that?


     [14] https://rawgit.com/w3c/imsc/tmichel07-patch-2/spec/ttml-imsc1-errata.html

   pal: Above is the link to the full rendered version.

   nigel: Is everyone happy now to merge that?

   group: no objections

   pal: I'll click Merge...


   nigel: Thanks Glenn for fixing the ToC issue and preparing a
   new draft for review.
   ... 3 things: 1. Can we publish the new WD? 2. How do we best
   work to allow the group to contribute to TTML2 spec?
   ... 3) Timing/planning for TTML2.


   <trackbot> action-463 -- Nigel Megitt to Add an issue on ttml2
   regarding 60 second minutes to point to the ttml1 one. -- due
   2016-04-28 -- PENDINGREVIEW


     [15] http://www.w3.org/AudioVideo/TT/tracker/actions/463

   close action-463

   <trackbot> Closed action-463.


   <trackbot> action-467 -- Thierry Michel to Ask glenn if a
   manual modification of the toc would be feasible to publish a
   new wd of ttml2 -- due 2016-05-05 -- PENDINGREVIEW


     [16] http://www.w3.org/AudioVideo/TT/tracker/actions/467

   close action-467

   <trackbot> Closed action-467.

   nigel: The WD we're proposing to publish is at

     [17] http://rawgit.com/w3c/ttml2/ttml2-WD-201605XX/spec/ttml2.html

   glenn: I've had 2 comments from Mike which I've addressed.
   ... He filed an issue on changing CEA to CTA, which I did.
   ... The open question on that point is that we had a variety of
   use of CEA in syntactic
   ... tokens, in named items. The question is should we make the
   change there as well.
   ... We don't have to decide right now. Where CEA appeared in
   the text elsewhere I changed
   ... it to CTA.

   mike: The other issue was that there was some text copied from
   a SMPTE spec, which we
   ... should either remove or seek permission for.

   glenn: I already replaced it.

   mike: That also resolves the checker error, which is quite

   pal: We had to train ourselves to ignore that error!

   glenn: The new pubrules require use of HTML5 - any obsolete
   HTML4 features are
   ... reported as an error and pubrules refuses to accept them. I
   had to make many changes
   ... in the content to remove those.

   plh: You're way ahead Glenn!

   glenn: We have a Changes document that we've been using since
   TTML2 that lists the
   ... cumulative differences between TTML1SE and TTML2. That
   document is listed in the
   ... Status section of the new proposed working draft.
   ... I also updated that particular document to use the new TOC
   format as well.

   nigel: I noticed that in §8.3.8 token was changed to NCName.
   Any particular reason?

   glenn: That was causing some parsing problems in something.
   ... This resolved a syntactic error report in the schema.
   ... If I recall NCName is a little more generic than a token
   (or the other way around)

   nigel: I think token is more generic.


     [18] https://www.w3.org/TR/xmlschema11-2/#NCName


     [19] https://www.w3.org/TR/xmlschema11-2/#token

   glenn: I have to check on what that was. We can change it back
   if it turns out to be a problem.

   PROPOSAL: Publish
   ml as a new WD of TTML2

     [20] http://rawgit.com/w3c/ttml2/ttml2-WD-201605XX/spec/ttml2.html

   RESOLUTION: Publish
   ml as a new WD of TTML2

     [21] http://rawgit.com/w3c/ttml2/ttml2-WD-201605XX/spec/ttml2.html

   plh: Now that we've agreed to associate a press release with
   this I'll need to liaise with
   ... the comm team to get a date for doing this - I can't do it
   in the next 15 minutes.

   nigel: okay, fine, let's hope that's pretty soon!

   pal: I'm available today to provide some additional text.

   nigel: There are two actions pending for the press release, so
   let's do those ASAP and push this along.
   ... Great, thanks all.
   ... The next thing to cover is process.
   ... There's a process that we use elsewhere which is to
   maintain a master branch as
   ... a proposed WD and have each issue resolved in a branch with
   a pull request that
   ... everyone can review.
   ... That's my preference, since it allows us to review work and
   publish WDs more frequently.
   ... It also makes it easier to track the changes for each

   glenn: This would create more impedance for me, so I'd rather
   not do it.

   plh: Putting changes into Pull Requests makes it easier for
   people to review substantive changes.
   ... This doesn't apply to typos or non-substantive changes.

   glenn: You can also use htmldiff or gitdiff. When we agreed to
   the process in Sapporo,
   ... in my mind this was just for IMSC and I thought I mentioned
   that I didn't intend to adopt
   ... it for TTML2.

   plh: It sounds like "it's my way or the highway", bluntly.

   glenn: I didn't say that.

   plh: That's what I'm hearing. We can change.

   nigel: The group has to review each change anyway prior to
   publishing a WD. Using PRs makes it easier.

   glenn: Using PRs introduces a time lag with dependencies
   between PRs that will give me major headaches.

   plh: We can have PRs on PRs. I've been using the PR process in
   a lot of groups and it really
   ... substantially increases the interactions. The process Glenn
   proposes does work but
   ... we are missing opportunities by sticking to it.

   glenn: It does dramatically increase the number of
   interactions, and they're overhead in my view.

   plh: Yes, that's the nature of this. We see a lot of benefits
   switching to github. It's not easy
   ... but we see a lot of benefits. We're not using github to its
   full potential yet.

   glenn: We're not using gh-pages either.

   pal: We used this for IMSC 1 to address all the comments for

   glenn: I think it's the editor's prerogative. I won't say no
   for ever. I might experiment
   ... and see if it works or not but I'm not committing to making
   that transition to full PR use.
   ... The tradition in W3C is that the Editor makes a lot of
   decisions about tools and processes
   ... and we don't want to get in the way of the Editors.

   plh: Yes, but we've been changing that.

   glenn: I'll give it a try and see what I think.
   ... There's nothing about the current process that prevents
   anyone else making a change either directly or by a PR.
   ... I'd prefer it with a PR because that makes it easier for me
   to review.

   nigel: There you have it! Can we have some symmetry?

   glenn: I've said all I can on this right now - I'll see what I
   can do and if I change my mind I'll let the group know.

   nigel: There's no notification on commit with github, but there
   is for pull requests.

   glenn: Can we switch that on?

   plh: Yes, we have tools to create notifications on commits,
   PRs, PRs with specific labels, all kinds of options.

   glenn: Are those configuration changes that I can make?

   plh: It depends on the kind of notifications you want. If you
   want the github built in
   ... mechanism we can do that, but otherwise we have other
   mechanisms that you cannot configure.

   glenn: Where would I go on the TTML2 repo to turn on
   notifications for commits?

   plh: It's under settings for the repository.

   pal: PRs also allows for discussion line by line about a
   specific change, which has been
   ... helpful in the past. You don't get that just with a
   notification. You can't attach comments
   ... to a notification of a change set.

   glenn: That's part of the increase in overhead that I was
   referring to. My preference is
   ... to get a bunch of suggestions and comments prior to each WD
   and then we can
   ... address those.


     [22] https://github.com/w3c/github-notify-ml-config/blob/master/mls.json

   plh: Is the proposal to notify on every commit?

   pal: No, I don't want that. I would like to be notified every
   time there's something
   ... substantive that requires group review.

   <plh> I see there "events": ["issues.opened", "issues.closed",

   nigel: By the way you can have a no issue Pull Request.
   ... Let's push on from this topic, with Glenn's offer to review
   this request on the table.

   <atai2> Need to leave...

   nigel: The last thing on TTML to cover is publication
   ... Pierre and I have come up with a revised timeline which is
   at [23]https://www.w3.org/wiki/TimedText/Publications
   ... [talks through the picture] Thanks to Pierre by the way for
   updating it.

     [23] https://www.w3.org/wiki/TimedText/Publications

   glenn: I like the goal of getting a CR out after TPAC. There
   are a number of substantial

   <plh> [24]https://www.w3.org/2016/09/TPAC/

     [24] https://www.w3.org/2016/09/TPAC/

   <plh> 19-23 September 2016

   glenn: ... items that need to be resolved, so that gives us
   around 4 months.

   nigel: Plus the discussion time in TPAC and any further
   resolutions needed.

   glenn: I think a lot of the details of the text and the
   features will be co-dependent on
   ... implementation activity around those features because we're
   progressively drilling
   ... down into semantics that have been by and large left in the
   past, in CSS etc and we're
   ... finding that we need to document those now.

   nigel: 2nd CR end of Feb 2017

   glenn: Okay, I'm comfortable with that as a timeline goal.

Profiles registry

   nigel: There are a number of issues open. I'd like to propose
   that Mike as Editor gets to
   ... merge PRs and close issues. Is that okay with everyone?

   plh: Yes.

   Mike: I'll close everything bar #8 and #13. I don't think there
   are any other outstanding PRs.

   nigel: Thanks Mike.

   mike: There's another step now. The Media Registration is
   complete and suitable to
   ... initiate the process with IANA. I'd like to get that kicked
   off so that codecs is confirmed
   ... or at least really close in IANA's eyes so that this can be
   used in MPEG.

   plh: I haven't looked at this and will add it to my todo list
   to figure out the next steps.

   <plh> [25]https://w3c.github.io/tt-profile-registry/

     [25] https://w3c.github.io/tt-profile-registry/

   mike: I'll close the issues, but the document should be
   available on the gh-pages branch

   plh: The master branch should be discarded.

   mike: Okay, from my perspective I don't care - let me know how
   you want to handle it.
   ... If the two are synched then I have no problem deleting

   glenn: Why do you need to delete the master branch to
   transition to gh-pages?

   plh: master and gh-pages are just names. Github lets you
   specify the default branch to
   ... use for a pull request. gh-pages has one advantage - they
   are automatically published
   ... by github.io and you don't have to use rawgit.
   ... Once you use that then having master is just confusing and
   could trip you badly.
   ... My recommendation is to delete master.

   glenn: I just want to check and confirm this - can you simply
   rename the master branch?

   plh: No, but it doesn't matter. Just create a branch with the
   new name and commit it.

   glenn: If I change the name of master locally and push it would
   that work?

   plh: I don't know to be honest. The way I do it is to create
   the new branch and then
   ... delete the unwanted branch.

   nigel: It's better not to use rawgit since that introduces an
   additional organisational dependency.

   pal: For previewing PRs you still have to use rawgit

   plh: That's correct.

   pal: That's where I think the most value is, since you want to
   review the pull requests.

   plh: Whichever branch you want to use is fine, but don't create
   the dependency to update both branches.

   mike: Back to profile registry, this is a normative reference
   from an MPEG document to
   ... the WG page. I don't know if there's a link anymore to the
   profile registry?

   nigel: We can create one if there isn't.

   mike: This needs to be discoverable from w3.org.

   <plh> [26]https://www.w3.org/wiki/TimedText/Publications

     [26] https://www.w3.org/wiki/TimedText/Publications

   plh: I'm guessing it's listed on the Publication Pages?

   nigel: No it's not.

   <scribe> ACTION: plh Add link to the TTML Profiles Registry
   page from the TTWG home page [recorded in

     [27] http://www.w3.org/2016/05/05-tt-minutes.html#action01]

   <trackbot> Created ACTION-468 - Add link to the ttml profiles
   registry page from the ttwg home page [on Philippe Le Hégaret -
   due 2016-05-12].

   mike: In the meantime people can keep working on the registry.
   We need a discoverable
   ... link to the registry document.
   ... The MPEG spec points to all sorts of codecs parameters for
   all sorts of media types
   ... and we require a link that can be used. It will be used by
   DASH and others for folk
   ... wanting to look to see what's in the files. I hesitate to
   point to a particular place because
   ... the document can move around.

   glenn: Can we fix this by putting it in the /TR space as a WG

   plh: Let me ask a question - what's the status of this? It's
   not a Rec.

   nigel: It's a WG Note. We do need to be able to update it.

   plh: That's fine, and it can be done automatically as well.

   mike: The issues on it are editorial. It's good to go today.

   glenn: If we have the document in the /TR space then we have a
   stable link that can
   ... be referenced in the ISO documents.

   PROPOSAL: Publish the TTML Profiles Registry as a WG Note

   nigel: On the basis that we can change it?

   plh: I need permission first time, then after that we can
   automate it.

   nigel: So we review a PR and then when we merge it into the
   default branch it gets
   ... published to /TR automatically?

   plh: That's correct.

   RESOLUTION: We will publish the TTML Profiles Registry as a WG
   Note on /TR and enable automated publication of updates
   following PR review and consensus.


   <trackbot> action-459 -- Thierry Michel to Process the PR at
   [28]https://github.com/w3c/tt-profile-registry/pull/3 and post
   for review to the reflector -- due 2016-04-21 -- OPEN

     [28] https://github.com/w3c/tt-profile-registry/pull/3


     [29] http://www.w3.org/AudioVideo/TT/tracker/actions/459

   close action-459

   <trackbot> Closed action-459.

   nigel: We did the two things in this action, not necessarily in
   the right order.

   <scribe> ACTION: plh Publish the TTML Profiles Registry as a
   Note and begin the IANA registry process [recorded in

     [30] http://www.w3.org/2016/05/05-tt-minutes.html#action02]

   <trackbot> Created ACTION-469 - Publish the ttml profiles
   registry as a note and begin the iana registry process [on
   Philippe Le Hégaret - due 2016-05-12].

   plh: I expect the Note will be published on Tuesday.
   ... You can keep doing whatever you want in the Repo Mike.

   nigel: Can we delete the master branch please?

   plh: Sure, I'm doing it now.

   mike: Then the only substantive thing to do is to close the
   open issues.

   plh: The github branch "master" has been deleted.


   nigel: I have had some feedback from members but have not
   collated it yet. The issue
   ... is how we deal with the clash with the Web & TV IG.

   plh: The schedule is draft at the moment.

   nigel: Yes, understood. Given the time I propose to come back
   to this next time.
   ... Thanks everyone for your flexibility today, and for a
   really productive meeting. [adjourns meeting]

Summary of Action Items

   [NEW] ACTION: plh Add link to the TTML Profiles Registry page
   from the TTWG home page [recorded in
   [NEW] ACTION: plh Publish the TTML Profiles Registry as a Note
   and begin the IANA registry process [recorded in

     [31] http://www.w3.org/2016/05/05-tt-minutes.html#action01
     [32] http://www.w3.org/2016/05/05-tt-minutes.html#action02

Summary of Resolutions

    1. [33]Publish
       ml as a new WD of TTML2
    2. [34]We will publish the TTML Profiles Registry as a WG Note
       on /TR and enable automated publication of updates
       following PR review and consensus.

   [End of minutes]

    Minutes formatted by David Booth's [35]scribe.perl version
    1.144 ([36]CVS log)
    $Date: 2016/05/05 15:58:45 $

     [35] http://dev.w3.org/cvsweb/~checkout~/2002/scribe/scribedoc.htm
     [36] http://dev.w3.org/cvsweb/2002/scribe/