Re: Re-chartering of the Second Screen Working Group

Hi Mark,

Thanks for the feedback, see inline.

On 09/09/2016 18:18, mark a. foltz wrote:
> Sorry for the late response Francois.  Overall scope and deliverables look good to me.
> Two questions on timelines:
>
> 1. The revised charter commits the group to deliver a FPWD of Presentation API Level 2 in Q2 2017.  I think it may be worth discussing at the F2F the level of interest in generating Level 2, and whether we have enough deferred feature work in the queue at the moment to merit immediate work.  The F2F may generate additional feature requests that make a stronger case.
>
> Also, one of the goals I would hope for Level 2 would be to gather feedback from Web developers on the V1 API and make improvements that benefit them.  Given implementation satus, the time frame might not be quite enough time to do that.  What is your experience with similar specs?

The "Level 2" spec in the charter was more meant as a way to give the group the freedom to work on such a spec than to commit it to deliver a FPWD at a certain point in time. The group may also decide not to work on a Level 2 spec after all if there's not enough interest or if the gap between the two levels is deemed too narrow at this stage. That's what I tried to convey in the "Expected completion" part. I updated it to be more upfront. It now reads:

[[
Progress on this specification depends on the adoption of the first version of the Presentation API and on the outcomes of discussions within the companion Second Screen Community Group. This Working Group does not expect to complete this specification by the end of this charter and may request to be re-chartered to finalize this work, possibly with a different scope.
]]

I also updated the description to mention "feedback from Web developers on the first version", and moved FPWD to Q3 2017. Note these milestones are estimates in any case and often not respected in practice for various (good or bad) reasons.

Does it look better?


>
> 2. For Presentation API V1, both PR and TR are scheduled for Q1 2017.  Is that realistic and in line with the time frame for similar specs?

Some specs stay in CR for ages, others just move forward. It all depends. I hope we can be among those that move forward :)

I think the timeline looks reasonable in our case, but I note it highly depends on implementation plans. I'm happy to update the milestones if you feel it's not realistic to expect implementations by end of February 2017.

 From a group perspective, to publish the spec as Proposed Recommendation, the group needs to address remaining issues, finalize the test suite, and create the implementation reports (one for controlling user agents, one for receiving user agents). Changes to algorithms could require another round in CR. That should not delay the spec too much unless we need to introduce more thorough changes to the API itself, which is hard to predict. There is work to do on the test suite, but that does not seem to be out of reach. The implementation reports are easy to assemble from test results... once the group has implementations to test!

Once published as a Proposed Recommendation, the AC gets asked for final approval to publish the spec as REC. That step is usually straightforward (there are plenty of occasions for members to raise concerns before that phase) and takes about 4 weeks.


Thanks,
Francois.

Received on Sunday, 11 September 2016 16:01:12 UTC