Re: Review comments on 3 March 2015 Process Editor's draft

On 2015-03-17 12:06, David Singer wrote:
>> On Mar 17, 2015, at 11:38 , Wayne Carr <wayne.carr@linux.intel.com> wrote:
>>
>>
>>
>> On 2015-03-16 21:09, Stephen Zilles wrote:
>>>> -----Original Message-----
>>>> From: Wayne Carr [mailto:wayne.carr@linux.intel.com]
>>>> Sent: Monday, March 16, 2015 4:10 PM
>>>> To: David Singer
>>>> Cc: Stephen Zilles; Charles McCathieNevile; W3C Process Community Group
>>>> Subject: Re: Review comments on 3 March 2015 Process Editor's draft
>>> [SZ] [SNIP] (and David Singer is ">>"
>>>>> I think if we can get to the point where groups never have expired charters,
>>>> it will be easy but irrelevant to say what is and is not allowed under those
>>>> circumstances.
>>>>
>>>> That's why I'd like it in the process that the WG closes when the charter
>>>> expires.  Seems there's no reason why it needs to expire rather than a short
>>>> extension.  I'd rather have it clear if the charter expires the WG is closed.
>>> [SZ] The problem with saying the WG is closed is that there are IPR considerations that are far from clear when a WG closes. We currently have a discussion as to whether re-chartering a WG creates a new WG or continues the existing WG. Any patent commitments that have been made clearly depend on the resolution of this discussion. The PSIG has been unable to resolve this issue, but adopting what you propose would IMHO push the conclusion in the wrong direction. I think that is why the AB has been looking at less drastic penalties for failing to re-charter.
> For me, it’s not that.  It’s that this is a common problem now, and if we get drastic about failing to re-charter BEFORE we get good at re-chartering, it’ll cause both panic and chaos. I’d prefer to avoid that.

It seems we can avoid the panic just by having many months before this 
get adopted -- or, additionally, by us asking the Director as part of 
the AC review to approve it to extend any WG this would apply to for 3 
months until the end of this year.  (and the Director could choose to 
keep automatically giving everyone an extra 3 months by extending.)

>
>> Oh no, not that again!!! :)
>>
>> I think that is an issue we have manufactured and we shouldn't let it keep us from doing what's sensible about recognizing that WGs that do not have an active Charter are not WGs.  We can word this so that it has nothing to do with the arcane issue of whether WGs are the same WG after recharter.  I'm not saying we should say that if a WG re-charters it is the same WG or a different WG - if a WG recharters it isn't in the situation I'm talking about at all.  I'm saying if a WG does not have an active Charter it is no WG at all. WGs have active charters - whether a series of different WGs or one WG that modifies its charter over time is irrelevant.
> I don’t think we disagree with you.  We just have to work out how to get there without causing panic and chaos.
>
>> The proposal is to add to Process section 6.2.8:  "If the Charter that a Working Group or Interest Group operates under expires, that group automatically closes and the Director MUST announce closure of the group to the Advisory Committee.”
> I would phrase as “is allowed to expire, i.e. is not either renewed or extended”.

There was an objection that closing the WG if the Charter is allowed to 
expire would somehow impact the debate on whether rechartering forms a 
new WG or whether the WG continues under a different charter (a debate I 
really don't get).  So I worded it without any reference to recharering 
or renewing.  I'll try again while trying to avoid implying anything 
about it being the same WG.

How about:

"If the Charter that a Working Group or Interest Group operates under is allowed to expire, that group automatically closes and the Director MUST announce closure of the group to the Advisory Committee. To avoid that, the charter could be extended or a modified charter with a later expiration could be approved.”








>
>> That could be avoided (trivially) by the Director simply granting a Charter extension.   All the Director has to do is say the Charter is extended and why.  If they re-charter it doesn't happen.  But, that says nothing about whether rechartering starts a new WG or not.
>>
>> 6.2.8 has provisions for the Director closing a WG for various reasons.  This proposal is a clarification where the group automatically closes if it does not have an active Charter to operate under and the Director just announces it is closed (the Director doesn't close it - it expires).
>>
>> Note that 6.2.8 says the Director "MAY close a group prior to the date specified in the charter in any of the following circumstances:"  Why can the Director no longer close a group AFTER the date specified in the charter?
> Saying that I may eat pasta for dinner, does not mean I may not eat it for lunch.  Where does it say he can’t?
>
>>   Is the intent that zombie groups without an active Charter are immune from action by the Director?  Probably not.  That restriction makes sense because without a valid charter the group is already closed so the Director can't close it! But, W3C has been allowing WGs to continue work after Charter expiration.  The SysApps WG continues at 5 months without a Charter and counting.  There are what amount to unfinished, abandoned specs in that WG and they're trapped there as long as that WG is considered as still open (without a Charter). Rules about not letting them publish, etc. do absolutely nothing about them trapping abandoned work.  It makes it so you can't even quit a dead WG because it may make decisions like dropping specs from it's non-existent Charter.
> We are in violent agreement that zombies must not happen.  We just need to manufacture the process and culture changes to get there.  And I’d rather have us get better at noticing the problem and dealing with it before it occurs, than have some very visible and embarrassing train wrecks cause people to get frightened.
>
>>
>>> Steve Zilles
> David Singer
> Manager, Software Standards, Apple Inc.
>
>
>

Received on Tuesday, 17 March 2015 22:36:21 UTC