Re: F2f agenda

Hi David,

To summarize, I do not think the group should approve the transition
to CR unless all required information is presented in its final and
complete form, i.e. not in a draft form in an email chain. Happy to
review the information when it is available, including over the
weekend.

To be sure, I do not have any technical comments on WebVTT, and have
no information that leads me to believe that, for instance, there are
features at-risk.

As I explained, I do not however see why this transition should be
treated any differently from other transitions, and the group asked to
consider incomplete information during a F2F that is already packed.

For instance, while I do not have an opinion on which change(s) are
substantive, a complete list of such changes needs to be presented
AFAIK.

>  how to report on implementations that are non-browser and hence cannot run any or all of the Web Platform Tests?

In the past, implementers have been able to self-report results,
outside of the Web Platform Test framework. I would think WebVTT could
use the same approach.

Best,

-- Pierre

On Fri, Jan 5, 2018 at 12:43 PM, David Singer <singer@apple.com> wrote:
>
>
>> On Jan 5, 2018, at 12:21 , Pierre-Anthony Lemieux <pal@sandflow.com> wrote:
>>
>> Hi David,
>>
>>> Hi, that’s what I thought I detailed in the previous email.
>>
>> In the past, substantive changes had to be provided as an attachment
>> to the specification. See [1] for instance.
>>
>> [1] https://www.w3.org/TR/ttml-imsc1.0.1/substantive-changes-summary.txt
>
> OK, that’s one way to do it.  That’s unusual though, most transition requests simply detail them in the transition request.
>
>>
>>> don’t think any other change raises to the level of ‘substantive’, they were clarifications.
>>
>> Even clarifications can be considered substantive per the W3C process [1].
>>
>> [2] https://www.w3.org/2017/Process-20170301/#substantive-change
>
> Yes, I know. I am the process CG chair. I am open to other suggestions as to what rises to the level of ‘substantive change’ here, though honestly, this is much more important for the PR transition (where it might affect IPR).
>
>>
>>> I thought that was also in the previous email. Implementation experience is documented by the Wiki, where we track implementations and their status.
>>
>> Will 2 implementations for every feature be required, as with other
>> TTWG specifications, or is the WebVTT group looking for something
>> different? This should be reflected in the SOTD of the document.
>
>>
>> Similarly, the SOTD should point to the implementation report.
>
> I know of no requirement that it be linked into the SOTD, but if the group would like it, we could look into it. I think it rather strange to link to such transient documents from a formal spec.
>
> There is a puzzle I am currently working on, as I said: how to report on implementations that are non-browser and hence cannot run any or all of the Web Platform Tests? If the group has any advice or suggestions, that would be helpful.
>
>>> The features at risk.
>>
>> AFAIK these need to be listed explicitly in the CR specification.
>
> I listed the candidates, but I don’t think that any pass the threshold; there are none that are worth removing in order to publish, because they represent important features. At the moment, we prefer to wait. There is nothing to link into the specification.
>
>>
>>> Where am I asking for anything different?
>>
>> In the past, the group considered a specification that included all
>> the information above (see [3] for a recent example), and where all
>> outstanding issues had been deferred by the group.
>
> The wide review issues are addressed or deferred; have a look at GitHub. Other issues and questions continue to come in and be addressed.
>
>>
>> [3] https://www.w3.org/TR/ttml-imsc1.0.1/
>>
>> Can you prepare an updated specification that matches these requirements?
>
> I think it is already there.
>
>>
>> Best,
>>
>> -- Pierre
>>
>> On Fri, Jan 5, 2018 at 12:06 PM, David Singer <singer@apple.com> wrote:
>>>
>>>
>>>> On Jan 5, 2018, at 11:33 , Pierre-Anthony Lemieux <pal@sandflow.com> wrote:
>>>>
>>>> Hi David,
>>>>
>>>>> Just a formal decision that the CR transition should happen.
>>>>
>>>> In the past, the TTWG asked for all process requirements to be met
>>>> before agreeing to a transition to CR, including boring stuff like:
>>>>
>>>> - documentation of substantive changes
>>>
>>> Hi, that’s what I thought I detailed in the previous email. The previous publication is the Wide Review WD, and the substantive changes are the addition of the color classes, and the definition of the conformance classes. I don’t think any other change raises to the level of ‘substantive’, they were clarifications.
>>>
>>>> - how adequate implementation experience will be demonstrated
>>>
>>> I thought that was also in the previous email. Implementation experience is documented by the Wiki, where we track implementations and their status. It lists the ones known as of the last update, and we expect to update it in the CR period.
>>>
>>>>
>>>> Why would it be different this time?
>>>
>>> Where am I asking for anything different?
>>>
>>>
>>>>
>>>> Best,
>>>>
>>>> -- Pierre
>>>>
>>>> On Fri, Jan 5, 2018 at 9:49 AM, David Singer <singer@mac.com> wrote:
>>>>>
>>>>>
>>>>>> On Jan 5, 2018, at 1:33 , Nigel Megitt <nigel.megitt@bbc.co.uk> wrote:
>>>>>>
>>>>>>> From my perspective, I think it is extremely unlikely that we will be able
>>>>>> to reach conclusions on those open WebVTT issues in time to agree to
>>>>>> transition WebVTT to CR during next week's F2F. I will do what I can in
>>>>>> the time available.
>>>>>
>>>>> To be clear, I am not expecting, or asking for, technical discussion of VTT issues at the Face to Face. Just a formal decision that the CR transition should happen.
>>>>>
>>>>> I would hope that the commenters can look at the large amount of (personal, volunteer) time that Silvia has spent on this on the last couple of months, and enable us to close the issues, but I recognize that time is always a scarce resource.
>>>>>
>>>>> I think that the major issue that needed addressing before CR was Andreas’ first wide-review comment on needing come built-in colors, and his linked second one on more closely defining conformance classes, both of which we have reached assent on. Nigel’s actionable comments that remain to be reviewed by him are appreciated, and are being addressed, but I am not sure that any need to hold up a resolution to proceed to CR; we expect to address them before CR (for the most part, pull requests are ready), and indeed I would suggest that we could ease pressure on Nigel by saying that we’ll take the proposed fixes into the CR and go ahead, and he need not review and comment if time is pressing (a problem I understand, as I have been trying not be be VTT co-chair for the longest time for the exact same reason).
>>>>>
>>>>> So, to be clear, we need the formal resolution from the WG that VTT is cooked (perhaps, rather over-done) to the CR point and we should formally declare CR and ask for implementations to catch up. No discussion of detailed issues (we’d need to be sure to have Silvia and ideally Simon and Philip, and Silvia is on vacation next week). That’s why I think that 30 minutes is generous.
>>>>>
>>>>> David Singer
>>>>>
>>>>> singer@mac.com
>>>>>
>>>
>>> David Singer
>>> Manager, Software Standards, Apple Inc.
>>>
>
> David Singer
> Manager, Software Standards, Apple Inc.
>

Received on Friday, 5 January 2018 21:34:48 UTC