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

Re: Evergreen Formal Objection handling (ESFO)

From: David Singer <singer@apple.com>
Date: Thu, 14 Mar 2019 08:20:21 -0700
Cc: Chris Wilson <cwilso@google.com>, W3C Process CG <public-w3process@w3.org>
Message-id: <FD30C858-59D2-42A3-A99D-AC7E6854CE9E@apple.com>
To: Philippe Le Hégaret <plh@w3.org>
I am completely baffled why you kept the 24 months. (A) we have no such requirement today on FOs and if we introduce it it should apply uniformly but, more importantly (B) how long the Director takes to issue a decision is not under the control of the WG and thus it’s inappropriate to put a constraint here that they cannot ensure.

On the marking, I said MUST on header and SHOULD on body, you made them both MUST. But imagine a draft written in English and using code snippets that represents decimal numbers throughout using decimal comma, not point. It is deeply confusing; is f(1,3, 4,6) a call with two or four arguments? I would formally object to a spec written this way, but we are not going to require the editor to mark every number as subject to FO. The point is that it is likely we will get FOs that are not about single, or a few, pieces of text, but about principles and styles that are spread all across it, so a MUST is not sustainable for every case, and in some cases, ridiculous amounts of work.

Lastly, you don’t distinguish the WG’s working draft from the W3C’s Evergreen Rec, and I think it important that we do. Most of the time they will be identical, but one can easily imagine a WG that decides to revert the ES after an FO is upheld, but keep a WD that has the problem still in it, in the new tech they want to introduce, while they work on modifying it to satisfy the decision. I don’t want a WD in that state to have no hint that there was an FO decision against it. Makes sense?

Sent from my iPad

> On Mar 14, 2019, at 8:05 AM, Philippe Le Hégaret <plh@w3.org> wrote:
> Here is my iteration on this, attempting to simplify David's proposal:
> [[
> * must document unresolved formal objections in the document header and relevant parts of the spec.
> * must ensure Director review of all pending formal objections before 24 months have elapsed.
> * must reflect resolved formal objections in the spec, in particular adjusting the document if the objection was subtained.
> EdNote: An alternative idea for FOs is to emphasize the role of the Chair in case of FO, ie the Chair should be more responsible for maintaining consensus.
> ]]
> 2 implications are:
> * one cannot publish if it doesn't reflect the Director's decision on FOs;
> * FO marking may be removed once the FO is resolved;
> * it does require that the relevant parts do note the FOs and not the document header (this is a change from David's proposal);
> I didn't touch the 24 months bits for now, understanding it is still unclear on how to change it.
> Regarding Chris' idea, while we all agree that the Chair is responsible for maintaining consensus, I believe we're trying to address the case where the Chair isn't able to, especially following an horizontal review. In other words, who is the final arbiter in those cases?
> Ideally, we would hope that the TAG could help but they are reluctant to add more on their plates.
> Philippe
Received on Thursday, 14 March 2019 15:20:50 UTC

This archive was generated by hypermail 2.4.0 : Friday, 17 January 2020 17:51:50 UTC