See also: IRC log
<trackbot> Date: 28 May 2015
<scribe> scribe: nigel
nigel: It's been a quiet couple of weeks. I'd like to discuss briefly the TTML2 Editorial Actions. AOB?
group: No AOB.
glenn: I've been doing a bit of work on Action-383 but have found no satisfactory conclusion to it yet.
action-383?
<trackbot> action-383 -- Glenn Adams to Missing XML declaration -- due 2015-04-30 -- OPEN
<trackbot> http://www.w3.org/AudioVideo/TT/tracker/actions/383
glenn: I'm still working on that one.
action-387?
<trackbot> action-387 -- Pierre-Anthony Lemieux to Collate list of potential implementors of imsc 1. -- due 2015-04-17 -- OPEN
<trackbot> http://www.w3.org/AudioVideo/TT/tracker/actions/387
pal: This is in progress. It's somewhat related to the plan for W3C to reach out to the implementors.
nigel: I think tmichel is back next week.
pal: Ok, I'll try to get this done and coordinate with tmichel. I want to avoid a lull after raising excitement levels with implementors.
action-396?
<trackbot> action-396 -- David Singer to Produce evidence of request for wide review for webvtt, for the archive -- due 2015-04-17 -- OPEN
<trackbot> http://www.w3.org/AudioVideo/TT/tracker/actions/396
courtney: dsinger sent some emails this week to try to wrap this up - he's working on it.
glenn: I expect to file some new
issues on TTML2 on the potential need for some additional
features
... related to handling corner cases with Ruby alignment. I
don't have the data to create those yet.
... Just to note the area that I'm looking at. There's a
probable need for adding some additional support
... for baseline alignment. Things get tricky with Ruby and
textEmphasis handling baseline alignment
... and generating even spacing for lineHeights whether there's
Ruby present or not. That's the area
... I'm referring to when I say we might need additional
features. Both CSS and XSL-FO have similar
... definitions of baseline properties that we can draw on for
doing this.
... The specific properties are three: "dominant" baseline
which is used to express the dominant baseline
... on the parent line area or on a child inline area. The next
one is "alignment" baseline which is used to
... express which baseline on the parent is to be used to align
the child. The third is baseline shift or
... baseline offset which is an additional movement of the
baseline related to the other two.
... Another thing I've been working on which is interesting is
ipd and bpd on a span. It turns out I've
... found a need for supporting struts, which are spans that
have zero size in ipd or bpd. Right now we
... effectively support them but don't say anything about them.
Simply by supporting ipd and bpd on
... span we support the semantics for struts but we don't talk
about them anywhere. It turns out they
... will probably be very useful for managing height of lines
in horizontal writing modes and width in
... vertical writing modes.
nigel: It sounds like a strut and a br are very similar things.
glenn: Not to me - br just
inserts a hard line break. Does a br introduce any vertical
space in a horizontal
... writing mode? It does based on lineHeight, if it follows a
previous br. The thing is that to alter the
... height of a line without having any content in it, there's
no way to do that today without using a strut.
... You could alter all the lines on a p by changing the
lineHeight property but you can't do it on a single
... line.
... We might also need to add to the condition system some new
layout function expressions that
... conditionalise application of styles based on layout states
for example line-first, line-odd, line-even,
... ruby-before, ruby-after etc. All these things have been
coming out of work I've been doing with edge
... cases on ruby recently.
nigel: I'm not aware of any other work on issues.
pal: I think the next step is to work with thierry and others to get input from non-W3C members on this.
nigel: Is there anything specifically that you think would help from W3C?
pal: As we discussed in Las
Vegas, this is a W3C spec but many of the implementors are not
W3 members.
... They may be a little concerned about sharing implementation
results with W3C, putting myself in their shoes.
... How will the results be used, why am I doing it, etc are
not immediately obvious. Since the process
... is contingent on those results we might be in a catch-22.
My recollection of the conclusion from the
... discussion in Vegas is it might make sense for W3C to reach
out to those implementors and explain
... why their input is valuable and how it would be used.
nigel: I'm fine with that - if
there are implementors who have concerns then we should be
understanding
... those concerns and trying to address them.
andreas: Do we have the idea of a webinar or something similar to explain IMSC to get more attention and explain more?
pal: I think somehow tmichel or maybe plh implied that they had ideas on what to do.
andreas: I think many people are
interested in that but maybe wouldn't join a bigger
conference.
... Half an hour or an hour to explain what's going on would be
good. Mike Dolan once did this for
... SMPTE-TT which was successful. If you can record it and
make it available later then it would be very helpful.
pal: +1 Let me take a note of
that for when I follow up with thierry.
... At least it should be a possibility we cover.
nigel: Has anyone looked at the ednotes spreadsheet and decided to volunteer?
pal: I've started looking at my one and have a question about it.
nigel: Me too for one of mine.
pal: I generated a sample image
for §10.2.10 disparity and glenn, are you wedded to the example
TTML?
... For instance the value of the disparity and the fact that
there's a single image?
glenn: Sure, go ahead and change
the example or mock-up to what you think is appropriate.
... You should have that flexibility.
pal: Thanks. Specifically, a
disparity of 0.7% is too small to see, and needs to be
increased, and also
... I'm thinking to have 2 regions, one with zero disparity and
the other with disparity so that you can
... easily see the difference between the two renderings.
glenn: Thank you.
nigel: My one was on system
model, §1.1 Figure 1. We know we want to change DFXP ->
TTML. But
... there are other things we might want to bring in.
... The first question is should profiles be in the system
model?
glenn: Interesting idea. The
system model right now focuses on data flow and types of data.
I would
... almost suggest a different diagram. If you mix profiles
then it might get too busy. I would suggest
... experimenting and seeing what makes sense.
nigel: My next question is does
it make sense to include transformation processors and
presentation
... processors in the system model, possibly relating the
content and processor profiles in there.
glenn: I would say yes, and also
including ISG generation as an option in the flow for
transcoding.
... We effectively have that informally in TTML1, though
without a concrete representation. From a reader's
... perspective, having an illustration somewhere that shows
something about that process visually
... would be useful. You could certainly add that into the
system model.
nigel: I'm also wondering if we
should explicitly name the flow from ISG into HTML/CSS to help
folk
... understand how TTML presentation fits into the web
context.
glenn: Yes, exactly.
nigel: Okay well I'll attempt to
do this and make it available for review.
... Going back to the spreadsheet, has anyone had a hard time
accessing it?
group: No (silently!)
nigel: Looking at the agenda, I
think there's no more to cover this week, so we'll draw to a
close.
... Just to remind everyone I'm still looking for a chair for
18th and 25th June.
pal: I'm happy to be available on standby for that.
nigel: Thank you very much
pierre!
... Thanks everyone, meeting same time next week. [adjourns
meeting]