W3C home > Mailing lists > Public > public-tt@w3.org > December 2008

Re: beginEnd002: par timeContainer and child with no duration

From: Silvia Pfeiffer <silviapfeiffer1@gmail.com>
Date: Tue, 16 Dec 2008 23:14:06 +1100
Message-ID: <2c0e02830812160414u393fe5d2u523bcd3677cc357f@mail.gmail.com>
To: "Geoff Freed" <geoff_freed@wgbh.org>
Cc: "Glenn A. Adams" <gadams@xfsi.com>, "public-tt@w3.org" <public-tt@w3.org>, "Sean Hayes" <Sean.Hayes@microsoft.com>

If ccPlayer is non-conformant, what is the problem about asking the
developers to fix the implementation? Excuse me for sounding naive and
having missed the history of TimedText and ccPlayer, but I would
assume that a non-conformant implementation should just be exposed as
such and be fixed.

Regards,
Silvia.

On Tue, Dec 16, 2008 at 10:49 PM, Geoff Freed <geoff_freed@wgbh.org> wrote:
>
>
> well, granted it's not the strongest of cases and would, as others have pointed out, lead to ambiguous timing situations, but i wanted to float the proposal anyhow.  ccplayer's behavior with regards to captions having begin with no dur or end may be problematic in the test cases, however.
>
> onward...
>
> g.
>
> ________________________________________
> From: Glenn A. Adams [gadams@xfsi.com]
> Sent: Monday, December 15, 2008 10:29 PM
> To: Silvia Pfeiffer; Geoff Freed
> Cc: Sean Hayes; public-tt@w3.org
> Subject: RE: beginEnd002: par timeContainer and child with no duration
>
> I also agree with Silvia, and would oppose the change suggested by
> Geoff. The proposed change would be a significant departure from SMIL
> timing semantics, which we have tried to maintain.
>
> Geoff, a better way for you to express what you want would be:
>
> <div timeContainer="seq">
> <p dur='1s'>This test counts from 0 to 10 in 10 seconds.<br/>0</p>
> <p dur='1s'>This test counts from 0 to 10 in 10 seconds.<br/>1</p>
> <p dur='1s'>This test counts from 0 to 10 in 10 seconds.<br/>2</p>
> <p dur='1s'>This test counts from 0 to 10 in 10 seconds.<br/>3</p>
> ...
> <p dur='1s'>This test counts from 0 to 10 in 10 seconds.<br/>10</p>
> <p end='10s'>This test is over.</p>
> </div>
>
> G.
>
>> -----Original Message-----
>> From: public-tt-request@w3.org [mailto:public-tt-request@w3.org] On
> Behalf Of Silvia
>> Pfeiffer
>> Sent: Tuesday, December 16, 2008 9:41 AM
>> To: Geoff Freed
>> Cc: Sean Hayes; public-tt@w3.org
>> Subject: Re: beginEnd002: par timeContainer and child with no duration
>>
>>
>> On Tue, Dec 16, 2008 at 11:51 AM, Geoff Freed <geoff_freed@wgbh.org>
> wrote:
>> >
>> >
>> > i agree with sean's explanation, as this is the expected behavior
> according to dfxp
>> now.  however, when we built ccplayer we implemented things a bit
> differently-- that is,
>> a caption that has a begin time but no end time or dur will display
> until the next
>> caption displays.  at that time, the first caption will erase just
> before the next
>> caption appears.  so in the case of this:
>> >
>> >      <p begin='1s'>This test counts from 0 to 10 in 10
> seconds.<br/>1</p>
>> >      <p begin='2s'>This test counts from 0 to 10 in 10
> seconds.<br/>2</p>
>> >      <p begin='3s'>This test counts from 0 to 10 in 10
> seconds.<br/>3</p>
>> >
>> > the first caption appears at 1s and is displayed until 2s, at which
> time it erases and
>> the second caption displays.  at 3s, the second caption erases and the
> third caption
>> displays.  etc., etc.  in dfxp terms, that equals this:
>> >
>> >      <p begin='1s' end='2s'>This test counts from 0 to 10 in 10
> seconds.<br/>1</p>
>> >      <p begin='2s' end='3s'>This test counts from 0 to 10 in 10
> seconds.<br/>2</p>
>> >      <p begin='3s' end='4s'>This test counts from 0 to 10 in 10
> seconds.<br/>3</p>
>> >
>> >
>> > we did things this way because caption software has not always
> depended on end times
>> to erase captions when the captions are timed to appear sequentially
> without pause.
>> >
>> > doing things the first way means less coding, which is convenient.
> i would lobby for
>> changing the spec, if it's not a big pain, to permit this behavior not
> only because it's
>> less work, but also because caption vendors will probably expect to be
> able to do things
>> this way.
>>
>> I respectfully disagree.
>>
>> To me, upon first reading the first example, it was clear that this
>> would add a new caption every second, but not remove any of the ones
>> before. This is a very convenient way of specifying a default end
>> value of "this last until the video ends whenever it ends". The best
>> means to support this is by not giving an end value and therefore
>> allowing it to last "forever". And it allows to have overlapping timed
>> text that lasts until the end.
>>
>> I think that if you wanted a text removed at a certain time, you'd
>> have to provide an end time. Otherwise you are open to all sorts of
>> misunderstandings. For example, what would you do with a specification
>> like this:
>>
>> <p begin='1s'>This test counts from 0 to 10 in 10 seconds.<br/>1</p>
>> <p begin='2s' end='4s'>This test counts from 0 to 10 in 10
> seconds.<br/>2</p>
>> <p begin='3s'>This test counts from 0 to 10 in 10 seconds.<br/>3</p>
>> <p begin='3s' end='4s'>This test counts from 0 to 10 in 10
> seconds.<br/>4</p>
>>
>> Would the third one appear at all?
>>
>> Best Regards,
>> Silvia.
>
>
>
Received on Tuesday, 16 December 2008 12:14:42 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 2 November 2009 22:41:39 GMT