Re: Proposed W3C Charter: Timed Text Working Group

98 views
Skip to first unread message

Henri Sivonen

unread,
May 3, 2012, 2:53:52 AM5/3/12
to dev-pl...@lists.mozilla.org
On Mon, Apr 30, 2012 at 7:30 PM, L. David Baron <dba...@dbaron.org> wrote:
> W3C is proposing a rechartering of the Timed Text Working Group.
> See:
> http://lists.w3.org/Archives/Public/public-new-work/2012Apr/0005.html
> http://www.w3.org/2012/02/timed-text-wg-charter
>
> Mozilla has the opportunity to send comments or objections through
> this Friday.
>
> It's not clear to me how this rechartering relates to the video
> captioning work in the HTML5 world.

IE 10 supports a tiny subset of TTML probably due to
Microsoft-internal politics or possibly out of anticipation of TTML
getting endorsed by a regulatory body. The subset seems to be so tiny
that it's almost a joke that it has anything to do with TTML. However,
a tiny subset of a larger spec carries the risk of the subset
ballooning into larger spec later on. I think all the previous reasons
against implementing the larger xmlns-laden XSL-FO-based spec still
apply even though Microsoft's subset probably isn't that harmful
unless it grows.

I think we should comment something negative about the continued
attempt to push TTML into HTML after Mozilla, Opera and Google have
shown a preference for WebVTT and even Microsoft has implemented
WebVTT and subsetted TTML to remove most of its features to make it a
mere syntactical alternative to WebVTT. In particular, I think we
should indicate we don't plan to implement TTML (assuming of course
that we still don't; I might have fallen out of the loop; it would be
prudent to check with Ralph Giles).

--
Henri Sivonen
hsiv...@iki.fi
http://hsivonen.iki.fi/

Ralph Giles

unread,
May 3, 2012, 10:51:47 AM5/3/12
to Henri Sivonen, dev-pl...@lists.mozilla.org
On 12-05-02 11:53 PM, Henri Sivonen wrote:

> I think we should comment something negative about the continued
> attempt to push TTML into HTML after Mozilla, Opera and Google have
> shown a preference for WebVTT and even Microsoft has implemented
> WebVTT and subsetted TTML to remove most of its features to make it a
> mere syntactical alternative to WebVTT. In particular, I think we
> should indicate we don't plan to implement TTML (assuming of course
> that we still don't; I might have fallen out of the loop; it would be
> prudent to check with Ralph Giles).

I'm in complete agreement.

-r

L. David Baron

unread,
May 3, 2012, 11:37:59 AM5/3/12
to Ralph Giles, Henri Sivonen, dev-pl...@lists.mozilla.org
So, then the question becomes, is said comment a:
(x) opposes this Charter and requests that this group be closed
[Formal Objection] (your details below).
or is it a:
(x) suggests changes to this Charter, and only supports the
proposal if the changes are adopted [Formal Objection] (your
details below).
I'm currently inclined towards the latter, I suppose, but not
strongly.

Does the following comment seem reasonable:
# We oppose the part of the charter that continues to try to push
# TTML into HTML5. At this point the market has settled on
# WebVTT: Mozilla, Opera and Google have shown a preference for
# WebVTT and Microsoft has implemented WebVTT and subsetted TTML
# to remove most of its features so that in their implementation
# it is merely a syntactic alternative to WebVTT. We certainly
# don't intend to implement TTML for HTML5 video.

Alternatively, if I choose the first option above, I could add:
# Given that captioning on the Web has clearly moved towards
# WebVTT rather than TTML, it's not clear why the W3C should be
# continuing to devote resources to TTML.

I'm inclined to cc: the response to the W3C Advisory Committee
mailing list.

-David

--
ğ„ž L. David Baron http://dbaron.org/ 𝄂
𝄢 Mozilla http://www.mozilla.org/ 𝄂

Ralph Giles

unread,
May 3, 2012, 12:15:04 PM5/3/12
to L. David Baron, Henri Sivonen, dev-pl...@lists.mozilla.org
On Thu May 3 08:37:59 2012, L. David Baron wrote:

> Does the following comment seem reasonable:
> # We oppose the part of the charter that continues to try to push
> # TTML into HTML5. At this point the market has settled on
> # WebVTT: Mozilla, Opera and Google have shown a preference for
> # WebVTT and Microsoft has implemented WebVTT and subsetted TTML
> # to remove most of its features so that in their implementation
> # it is merely a syntactic alternative to WebVTT. We certainly
> # don't intend to implement TTML for HTML5 video.

It does. As far as I can tell the charter consists *only* of
developping a new TTML spec.

Note there are already two community groups at the W3C, one pursuing
WebVTT[1] and one working on TTML[2]. We participate in the former,
which has been doing useful work to develop the spec. The later as been
much less active, at least in public.

> Alternatively, if I choose the first option above, I could add:
> # Given that captioning on the Web has clearly moved towards
> # WebVTT rather than TTML, it's not clear why the W3C should be
> # continuing to devote resources to TTML.

That is also a reasonable response. Producing an improved TTML spec is
a fine goal, but not something I see a need for in the domain of the
web. I would also say the current process, with the community group
developing a community WebVTT spec, is working.

I don't have a strong feeling either. If we think a more formal working
group would help with buy-in from other organizations, we should
support the creation of timed text working group focussed on improving
the WebVTT spec. We supported the WebVTT community group in the hope
that it would be more accessible to large implementors than the earlier
whatwg process, and the same argument applies here.

-r

L. David Baron

unread,
May 3, 2012, 1:09:36 PM5/3/12
to Ralph Giles, Henri Sivonen, dev-pl...@lists.mozilla.org
I took an in-between tack in my response:

# We oppose the part of the charter that continues to try to push
# TTML into HTML5. At this point the market has settled on
# WebVTT: Mozilla, Opera and Google have shown a preference for
# WebVTT and Microsoft has implemented WebVTT and subsetted TTML
# to remove most of its features so that in their implementation
# it is merely a syntactic alternative to WebVTT. We certainly
# don't intend to implement TTML for HTML5 video.
#
# Given that captioning on the Web has clearly moved towards
# WebVTT rather than TTML, it's not clear why the W3C should be
# continuing to devote resources to TTML. This makes me almost
# inclined to choose the "opposes this Charter and requests

and I'll see what the reaction is and how to go from there.

L. David Baron

unread,
May 14, 2012, 5:17:18 AM5/14/12
to Ralph Giles, Henri Sivonen, dev-pl...@lists.mozilla.org
On Thursday 2012-05-03 07:51 -0700, Ralph Giles wrote:
> On 12-05-02 11:53 PM, Henri Sivonen wrote:
> > I think we should comment something negative about the continued
> > attempt to push TTML into HTML after Mozilla, Opera and Google have
> > shown a preference for WebVTT and even Microsoft has implemented
> > WebVTT and subsetted TTML to remove most of its features to make it a
> > mere syntactical alternative to WebVTT. In particular, I think we
> > should indicate we don't plan to implement TTML (assuming of course
> > that we still don't; I might have fallen out of the loop; it would be
> > prudent to check with Ralph Giles).
>
> I'm in complete agreement.

So I'm hearing a bunch of pushback on this.

My understanding is that the main motivation for wanting something
TTML-based is that there are huge amounts of content in it or in
formats that have existing mappings into it (in particular, movie
and TV content in the libraries of the producers of that content).

But it sounds like there's not pressure for all of TTML -- the
interest in really in SMPTE Timed Text (a significant subsetting of
TTML) or some other slight variants of SMPTE, and it sounds like
there might be willingness to subset further if that would help
(based on what features the existing pools of content want to use).
(In other words, to build TTML 2.0 on top of the subset of TTML 1.0
rather than the whole thing.)

It also sounds like there may be willingness to move TTML away from
XSL-FO towards HTML+CSS.

What do you think of the tradeoff of a compromise TTML like this
versus WebVTT? And if you still think we should advocate for WebVTT
given that situation, what's the rationale we should give?

L. David Baron

unread,
May 15, 2012, 10:45:10 AM5/15/12
to dev-pl...@lists.mozilla.org
On Monday 2012-04-30 12:30 -0400, L. David Baron wrote:
> W3C is proposing a rechartering of the Timed Text Working Group.
> See:
> http://lists.w3.org/Archives/Public/public-new-work/2012Apr/0005.html
> http://www.w3.org/2012/02/timed-text-wg-charter

There's now also a proposed set of revisions to this charter at:
http://www.w3.org/2012/05/timed-text-wg-msft.html

The comment deadline has been extended to this Friday (May 18).
Reply all
Reply to author
Forward
0 new messages