W3C home > Mailing lists > Public > public-w3process@w3.org > September 2019

Re: Continuous Development Process TPAC Slides

From: Florian Rivoal <florian@rivoal.net>
Date: Fri, 6 Sep 2019 13:05:19 +0900
Cc: W3C Process Community Group <public-w3process@w3.org>, Philippe Le Hegaret <plh@w3.org>
Message-Id: <0001D8E9-E0ED-48B2-BF07-C32D11184753@rivoal.net>
To: David Singer <singer@apple.com>

> On Sep 6, 2019, at 7:49, David Singer <singer@apple.com> wrote:
> Hi
> Some comments

Some comments on the comments (keeping in mind I didn't write the slides)

> I thought we had agreed to separate Registries as a separate, simpler, case, and we’re presenting them separately. Yes, I think the “continuous review” model used by Living Standards (notifications on issues and changes, and the ability to comment) is the right model also for Registries, but even then, Registries are simpler — being atomic, it’s much easier to back out a specific change. So many of the mentions of Registries should be in the Registries report.

I don't think this presentation should spend much time on registries. The "@@See separate AC presentation" should be replaced with a link to https://www.w3.org/2019/Talks/TPAC/ac-registries/

I have no strong opinion on whether the rest of the slide should be removed or trimmed down or left as is, but that's certainly not where this presentation should spend its time.

> I still don’t understand why getting a Contribution License from WG members fixes anything in the W3C context, where WG members grant a full-spec. license.

There can be fairly long stretches of time between the kinds of drafts that trigger patent exclusions/commitments. Contribution licenses give some coverage in the meanwhile. I don't think it's that critical, but it's a nice to have improvement, both on its own merits, and to bring us to parity with the WHATWG. 

> The title says “continuous standard development” and so one might expect simple Living Standards, i.e. Evergreen, to appear. But Slide 3 immediately diverts into improving the Rec Track, i.e. Everblue. Overall, I think this slide deck does not present simple Living Standards well or as a reasonable direction and choice. Instead it presents Improved Rec. Track as the primary effort and Evergreen as some sort of half-thought-out dead alternative.


> The Design Intentions does not say what the AB has repeatedly said: that a simple Living Standards process and improving the Rec. Track are not in opposition, and we could do either or both.
> The Proposal Part is only a proposal for Rec. Track improvements, not for simple Living Standards.

The first part of the proposal is for REC track improvements that we want to do anyway AND that get us to (or closer to, depending on opinions) continuous spec development: it enables keeping TR continuously up to date at all stages of spec evolution, without blocking on the Director (or anyone's) review. But it does keep checks and balances and reviews when you want to claim the next level of maturity (after which you can continue to keep things up to date at that level of maturity). And you can incrementally add features if you want to.

To me, that's the core of continuous standards development, and I'd be satisfied with that. Not all have to agree with me, and some may want something with fewer checks (and correspondingly less Process). There's the evergreen approach for that.

We should not do is ask "do you want evergreen, or do you want the to fix the REC track", as there's no conflict between the two. What we should ask is "given that we're fixing the REC track in this way, does that satisfy those who were previously frustrated with it, or do we nonetheless need Evergreen?". Which is by and large how I read this presentation.

> Then we get to the “Additional alternate track?” slides, which says “We are not supporting”.  I have no idea who “we” is. It also implies it’s dead (“originally tried”).

we = the majority of voices in the AB last time we talked about this?

Evergreen is presented, but the last couple of AB meetings showed it to be increasingly in disfavor, and an increasing number of people thinking that the proposed fixes to the REC track (everblue+everteal) got us what we were looking for, without sacrificing what we wanted to keep. But it is mentioned nonetheless, since that's a plausible approach that we have worked on, and the AC should hear about it.

I think we could tweak wording and emphasis, but I think the overall structure is about right.

Received on Friday, 6 September 2019 04:05:48 UTC

This archive was generated by hypermail 2.4.0 : Friday, 17 January 2020 17:51:52 UTC