Re: 2014 Process Regresses Editorial Revision of RECs

On 2014-10-15 17:16, fantasai wrote:
> On 10/15/2014 03:57 PM, Wayne Carr wrote:
>> fantasai wrote:
>>> Since I don't see any reason to make this more complicated than it
>>> was in the 2014 Process, I would like to see this change reverted.
>>
>> This is supposed to be "Editorial changes or clarifications that do 
>> not change the technical content of the specification."
>>
>> It needs some sort of review to make sure it really is just editorial 
>> and also that it isn't making a change people think
>> makes it worse.  One possibility could be Director's approval and the 
>> AC can choose to appeal the Director's decision with the
>> usual timeline (3 weeks to appeal). The problem is you don't want 
>> something published that accidentally causes new licensing
>> commitments when you think you aren't.
>>
>> A Director's decision plus appeal rather than PR doesn't really save 
>> that much time (3 weeks during which an appeal can happen
>> vs 4 weeks review).
>>
>> Another alternative could be a patent policy change so that Edited 
>> Recommendations do not create any additional Essential
>> Claims (even when someone accidentally adds something that otherwise 
>> would create new Essential Claims).  And make them always
>> be called Edited Recommendations so that applies.  Then we could 
>> loosen up the rules on how easy it is to change the text
>> without worrying it is changing patent commitments.  That probably is 
>> worth doing anyway because these aren't supposed to be
>> causing patent commitments.  It also simplifies doing them when there 
>> isn't a WG to do it.
>
> I think it would be important for a WG to explicitly review and
> ratify each change that goes into a REC, including editorial ones,
> and not just rely on the editor to make that judgement--to ensure,
> as you say, that the change is indeed merely editorial and is an
> improvement.
>
> But cycling through PR is not likely to help anyone. Neither the
> AC nor the Director have any particular expertise that helps to
> notice that the change is not editorial. You may as well ask
> Random Technical Mailing List Dude for the official sanction.

Even if that were the case, they do have co-workers who do have 
expertise.  Typically, the AC is more likely to be involved than WG reps 
when something impacts licensing commitments or represents the W3C 
Member's organizations view (like the AC Review of Proposed 
Recommendations or WG Charters).

> Plus, if someone screws up, people should be able to raise that
> as a problem 6 months later, not just within the 4-week PR period,
> and have that corrected immediately.

Of course people could raise new errata later.  But, once its out as a 
REC there are associated licensing commitments.  The entire spec could 
be rescinded later, but that's a drastic action and even that wouldn't 
undo everything.  A PR period allows a look before it gets published, as 
would a Director's decision with the possibility of an AC appeal.

That's why changing the patent policy so Edited RECs don't add essential 
claims would ease the problem because if mistakes were made and changes 
actually were substantive, it would not impact licensing commitments.

>
> Requiring a cycle through PR for this is not an improvement, it's
> a regression.
>
> If you're troubled about the possibility of substantive changes
> masquerading as editorial ones, let's put in the Process that
> at least 3 people, at least two of whom are members of the WG,
> at least two of whom are not editors of the spec in question,
> and each of whom are employed by different companies, have
> vetted the exact wording of a purportedly-editorial change
> before it can be published, and it is the responsibility of the
> chairs to ensure that this has happened.* That will do a much
> better job of guarding against your concerns, with much less
> useless overhead, than cycling through PR.

For the reasons, above that wouldn't work.

What I was trying to point out in my initial note was one reason why it 
may have been changed, and why it may be a positive change.

It would be good to be able to make these non-substantive changes even 
if the WG wasn't around.  If the patent policy changed so, they were 
guaranteed not to impact licensing W3C staff or some special WG could do 
maintenance like that when there wasn't a WG.

>
> * Alternately, the chair can certify that this was merely a
> typo and doesn't need such review.

The Director/appeal variant I suggested would be the WG would ask to 
republish and say its a typo.  The director says yes, its a typo and 3 
weeks later it gets published.  (or same thing if its rewriting a 
paragraph because it was confusing).

>
> ~fantasai
>
>
>

Received on Thursday, 16 October 2014 02:44:59 UTC