Timed Text Working Group Teleconference

20 Feb 2009

See also: IRC log


Franz, Sean, Plh, David, Geoff
Glenn (at sea), Andrew (vacation), John


Test Results

plh: Content tests: would be good to know if Abode and WGBH are planning to support nested div and spans. That's a major set back for our tests, especially the timing ones.
... NCAM doesn't support xml:space. Nothing to be done there. We have enough implementation to move forward for it.
... timing tests fail because of nested div, and lack of timing model support in Adobe and NCAM. That's a big problem for us.
... I don't have Glenn implementation but I expect it'll be the best of all, but I'm uncomfortable to use my own HMLT5/JS implementation to move the draft forward since I see my implementation as experimental/proof of concept, not meant to become a product or anything beyond that.

Sean: Proof of concept is good enough. I'm focusing on the timing part for my implementation so I expect good news on that from Microsoft, if time allows. Not sure yet to what extend I'll implement the styling section. Might still be on target for end of the month.

David: I have Glenn implementation from 18 months. I'll send it to plh. Can you have test results for it in 2 weeks?

plh: I could, even if Timed Text isn't one of my priorities at the moment. Testing standalone players require a lot more time since I don't have the advantage of the framework.

David: what should be our focus in the upcoming weeks?

plh: I think that developers, in particular Adobe, WGBH and Microsoft, should look at the test results, starting with the failed ones. Check that the test is valid, check that my reported result is correct, and report back to the Group whether they plan to pass the test or not in the upcoming months. To my surprise, Andrew reported that his implementation can be improved when we talked about dynamicFlow. We need to understand what is planned to be supported by end of this summer at the latest. Our basic profile depends on that knowledge since we could keep more features that are not supported currently if we get implementations by then. I'll run the tests against Glenn's implementation and we could start by eliminating any feature that is not supported by Glenn for our basic profile (like outline?).

[Goeff arrives]

Goeff: further development on the NCAM implementation depends on fundings, which we don't have at the moment unfortunately. I'll check whether we could fix our span element bug.

plh: alpha version is good enough for our test report.

plh: we're doing great on the test coverage. Only missing one is dynamicFlow. No need to spend more cycles on that for now. We still need to figure out what to do for our transformation profile during the CR phase. I still believe that demonstrating transformation between existing formats would be great. Not looking for round-tripping but at least transformations.

Action items review

Plh: Didn't do ACTION-17. Priority is getting test results.

Plh: re ACTION-20. It seems that this action is superceded but not sure

Sean: I believe I took care of section 11 (Animation) tests.

plh: will check state of ACTION-20 and might close it if we're good.


plh: Didn't we agree to keep dynamicFlow?

Goeff: yes, but we added rollup behavior.

ACTION: Glenn to add the rollup value to dynamicFlow

ISSUE-1 is closed


Sean: I've been using the naration role but we probably to separate those two out [...]

David: can't see having any problem in adding more values to role

plh: we need an email from Sean listing the new value(s) and their description(s), then we can action Glenn.


plh: we declared the handling of xml:lang in the NCAM implementation incorrect. I don't think there is anything that should be done in the spec at this point.

Goeff: ok.

ISSUE-14 is closed


Sean: css 3 changed their alpha value definition
... in our spec, it's still a very large number of values
... I was trying to constrain that a little to align with others
... don't remember how the discussion ended up
... Glenn didn't agree to put the proposed type into the xml schema

plh: Did Glenn agree to change the schema?

Sean: I don't think so. We need Glenn to discuss this


Sean: you can have style elements as child of region
... that's an other way to apply to the region
... it isn't referential
... but that's a third way of doing it, and it strictly equivalent to applying the inline style to the region
... the algorithm presented in 8.4 doesn't cover that case. So we would need to extend the algorithm
... there is a certain of work either way
... do we want to keep the style element or not?

Goeff: I found it useful

ACTION: Glenn to add support for style element inside region in section 8.4


Sean: my preference is to have the default as par for the body element

Goeff & Plh: +1

Plh: I'll figure out if it is the case in the spec and will follow up with Glenn if not

Next meeting

plh: we're good on the issues front for now. we're good on tests. Do we need a meeting next week since we don't expect a lot of progress on test results?

David & Sean: No meeting next week. We need to concentrate on the test results between now and March 6.

Regrets from Sean on March 6. Regrets from Goeff on March 20.

Next meeting is on March 6.

[End of minutes]

Minutes formatted by David Booth's scribe.perl version 1.133 (CVS log)
$Date: 2009/02/20 19:14:16 $