Re: Publishing WebVTT as a Proposed Recommendation (CfC)

My high level review question here is: are all the features implemented?
This matches the intent of the CR exit criteria.

## Regions and positioning

In our call on Thursday I satisfied myself that the region and positioning
test failures weren’t a reason to block progression, because the use of
lines with numbers in the int number space is supported and doesn’t cause
any particular problems, and the feature _is_ implemented in a way that
would work with those numbers. Gary took the action to create tests that
verify that.

* I would like positive confirmation that the int number space tests for
lines do indeed pass on at least two implementations before we go ahead
and request PR publication.


## Text wrap: balance

We did not discuss this feature on Thursday.

The initialisation of the text-wrap CSS property to the value “balance" is
a MUST requirement in the specification (§7.4). However the implementation
report shows the relevant test is failing because:

> This is failing because `text-wrap: balance;` isn't available in any
>browser.

This is of some concern in itself; it is even more concerning that this is
a normative requirement and it is dependent on the definition in CSS Text
Module Level 4, which is at WD. caniuse.com doesn’t report any results for
text-wrap. Web Platform Tests does not appear to include any tests for CSS
text -> text-wrap.

My assessment of this is that we have a mandatory feature for which there
is no suitable normative reference available and no evidence of it ever
being implemented. 

Proceeding with this as-is would surely be against the spirit of the CR
exit criteria; I’m struggling to imagine how we could explain it in any
reasonable way.

* I think the only action to take is to mark the use of text-wrap: balance
as at-risk and remove it.

Nigel


On 06/06/2019, 20:48, "Philippe Le Hégaret" <plh@w3.org> wrote:

>
>
>On 6/6/2019 12:42 PM, Nigel Megitt wrote:
>> Please can you extend the deadline to June 20 in line with the group¹s
>> decision policy [1]?
>> 
>> [1] https://www.w3.org/2018/05/timed-text-charter.html#decisions
>> 
>> This allows for 10 _working_ days, which we normally take to mean 2
>> calendar weeks.
>
>Of course. s/June 17/June 20/ it is (it won't make much of a difference
>at the end anyway for the Director).
>
>Philippe
>
>> Thanks,
>> 
>> Nigel
>> 
>> 
>> On 06/06/2019, 16:53, "Philippe Le Hégaret" <plh@w3.org> wrote:
>> 
>>> Based on the current issues, the implementation report, and today's
>>>call:
>>>
>>> https://github.com/w3c/webvtt/issues
>>> https://www.w3.org/wiki/TimedText/WebVTT_Implementation_Report
>>> https://www.w3.org/2019/06/06-tt-minutes.html#item02
>>>
>>> The document without the features is accessible from the following PR:
>>>   https://github.com/w3c/webvtt/pull/453
>>>
>>> Gary intends to raise some issues related to some of the corner cases,
>>> like clamping unsigned long values, but not a blocker at this point.
>>>
>>> If anyone objects to publish WebVTT as a Proposed Recommendaiton,
>>>please
>>> speak up no later than June 17.
>>>
>>> Thank you,
>>>
>>> Philippe
>>

Received on Monday, 10 June 2019 10:23:13 UTC