Re: w3process-ACTION-47: Produce a proposal for addressing wayne's "comment 9" - allowing appeal where the director's decision isn't the same as the proposal sent for review.

On 2015-03-18 16:30, David Singer wrote:
> Overall, the whole question of appeals badly needs a cleanup.  In some places it says an AC Rep can initiate an appeal, in some other it says the AC (but I think it means the Rep.).  It’s clear that 5% of the AC have to agree to the appeal to cause a vote (within a week), but then is silent on whether there is a quorum requirement for the vote itself, and what the passage requirement is (50%? of those voting?).  How long is the voting period?  And so on.
>> On Mar 18, 2015, at 14:14 , Wayne Carr <> wrote:
>> Alternate text to correct the problem:
>> Current: "When Advisory Committee review immediately precedes a decision, Advisory Committee representatives MAY only appeal when there is dissent. "
>> Proposed: "When Advisory Committee review immediately precedes a decision and the Director approves the proposal, Advisory Committee representatives MAY only appeal when there is dissent. “
> so you want it to say that there can be appeal if either (a) there was dissent in the AC vote or (b) the Director’s decision is contrary to the unanimous (among those voting) opinion of the AC?


> like I say, I get what you are saying, but has that ever happened? will it?  The Director already knew that there was a unanimous position, and went the other way.  Why would he do that, if a single appeal could trigger a vote that might override him?

WG Charter review.  Director thought it a good idea when submitted. In 
AC Review, no formal objections (so no dissent), but AC reps asked for 
changes while not requiring them.  Director decides the charter is a 
mess and rejects it.  Some AC rep thinks the charter was fine and 
shouldn't have been rejected.  5% agree.  Vote happens and it passes.

That doesn't seem impossible. It doesn't require bad behavior by 
anyone.  It's all reasonable.  This is a very simple change.  What harm 
could it do?  And if it never happens what harm could it do to fix what 
clearly is an error?

> If we don’t make this change, what other course does the community have?
> Let’s say the AC, in a rare display of unity, REJECTS something. The Director then says ‘yes’, despite the entire community saying ‘heck, no’ (or at least the 3 who voted). Seems very unlikely.

In that case, we already can appeal, because someone did a formal objection.

>   Let’s say the AC accepts something.  The Director says ‘no’, despite the entire community (well, both voting) saying ‘yes’. Can’t they just ask again (e.g. ask the chair to make a trivial change to the document and re-request the transition)?

You can't get back to the AC review unless the Director wants to. Make 
it a WG charter like above.  Everyone says yes, Director says no because 
of something that came up in the Review.  Only the Director can choose 
to submit a Charter to the AC.  When the review and decision are over, 
there's no way for the AC to cause it to be reviewed again.  Appeal 
right then  after an AC Review is all we can do.

>> The text is trying to get rid of the case where the director approves, no one disagreed in the AC review and then an AC rep appeals because, while they didn't submit a review, they later decide they don't like it.  But, it also removed the case where there was no dissent because everyone in the AC approved, but the director rejected it.
>> Not a big deal, but why needlessly have this wrong when there is a trivial wording fix?
> The wording is not the hard part…it’s the likelihood.

I'm hoping the Charter one is more convincing.  But, also -- it clearly 
was intended to prevent appeal when the AC Review and Director already 
completely agreed to approve - something good to prevent.  But, the 
wording also banned appeal when the AC all say YES and Director says NO.

>>   But do we want:
>> - AC cannot appeal a rejection when every AC rep who submitted a review disagrees with the Director
>> - AC can appeal a rejection when  every AC rep who submitted a review agrees with the Director
> no, there has to be dissent.  Some on one side, some the other.
>>>> I'd think no one intends for that to be the case.  It's an error due to wording.  What the wording is trying to say is that the AC cannot appeal in the following case:
>>>> Example 2:  AC Review on publishing a Recommendation.  400 AC reps approve publishing.  0 formal objections.  The Director approves publication.  The AC cannot appeal.  No AC member can ask to override if there were no formal objections in the AC review.  That's good, but they worded it so it also bans appeal if the Director rejects publishing when no one objects to publishing!
>>> I think that may be intentional.
>> right, it's a wording error.
> I said it may be *intentional*.  The Director can say “you goofballs all think this is a good idea, but I am The Director!” This is like a presidential veto.

If it was intentional then he would be saying:  "400 yes and 0 no and I 
can say NO anyway and nothing you can do about it!  But, 399 yes and 1 
NO and then I can't say NO because I'll let you appeal in that case."  
That doesn't sound intentional.

Some AC rep can decide to always object in a trivial way to everything 
and then the AC always has the right to appeal the results of AC Reviews 
(of the things on the appeals list).  That would be silly - the point is 
the change isn't doing something we couldn't already do in a silly way 
and it corrects in a simple way what looks like unintentionally bad wording.

> For all the principles, the W3C is formally a Monarchy (Dictatorship, whatever): we only ‘advise’…

Not with the appeal process ... it's a system that allows the Director 
to help keep things going with flexibility and not having to rely on 
numerical votes in order to try to do the best thing. But, if we really 
object, we can appeal and then the AC does decide.  It never has 
happened, so it must work fairly well.

There are some things where it is a monarchy.  Proposing a Charter to 
start a WG for instance.  Or resubmitting a Charter.

But, this requested change isn't about changing that balance at all.  
It's a mistake.  Changes a case that either will never happen or if it 
did -- we really wouldn't want to happen.  So there is no harm in 
changing it and we should.

> David Singer
> Manager, Software Standards, Apple Inc.

Received on Thursday, 19 March 2015 04:22:37 UTC