RE: editing RECs after publication -> RE: "Living Standards"

Some suggestions to get the TR publications to be closer to the most recent WG draft:

First Public Working Draft and Last Call Draft need to be fixed because of the patent exclusion period associated with them. (But they can be followed at any time by a regular Working Draft.)

Candidate if it has a substantive conformance changing edit has to go back to Last Call for the exclusion review.  But for edits that don't change conformance, why not allow edits without issuing a different Last Call?  (but carefully indicating what changed after first publication - or even show the edits if the WG wants to). 

Proposed Rec - same as Candidate Rec for non-conformance changing edits, except frozen during AC review where PR has to be stable.

But for regular heartbeat working drafts publications in the TR list, why can't they be updated as often as the WG wants to?
(so after the WG publishes a Last Call or Candidate Rec or Proposed Rec, they can keep  publishing a regular Working Draft update to the TR list as often as they choose - like every 2-3 weeks).
Wouldn't that solve the problem of the most recent TR publication being far out of date?  Why not make it very simple to publish a WD every few weeks?

And also mark previous TR publications as superseded if there is a later TR publication.  (superseded may be a better word than obsolete.  It could be for some specs, older RECs are still used in some setting after they have been superseded).

>-----Original Message-----
>From: Carr, Wayne [mailto:wayne.carr@intel.com]
>Sent: Thursday, February 09, 2012 4:26 PM
>To: Robin Berjon
>Cc: Charles McCathieNevile; public-w3process@w3.org
>Subject: RE: editing RECs after publication -> RE: "Living Standards"
>
>+1
>
>>-----Original Message-----
>>From: Robin Berjon [mailto:robin@berjon.com]
>>Sent: Thursday, February 09, 2012 2:08 PM
>>To: Carr, Wayne
>>Cc: Charles McCathieNevile; public-w3process@w3.org
>>Subject: Re: editing RECs after publication -> RE: "Living Standards"
>>
>>On Feb 9, 2012, at 20:49 , Carr, Wayne wrote:
>>>> 2. It would be good if W3C process allowed for simple editing of "finished"
>>specs.
>>>> As far as I know it is really easy for a WG to approve errata, which
>>>> are meant to be linked from a spec anyway, although there is no
>>>> mechanism for a spec to say "there are *actual* erratat there you
>>>> should look at" as opposed to "there might be something...". I've
>>>> never tried to push through a Proposed Edited Recommendation
>>>> (although I have added work for people who did try to do so by
>>>> asking for it to
>>reflect reality better, which they kindly did).
>>>
>>> http://www.w3.org/2005/10/Process-20051014/tr.html#rec-modify
>>> (section of the process doc)
>>>
>>> For changes that don't impact conformance (like changing examples, or
>>> simple
>>clarifications), the process says: "The first two classes of change
>>require no technical review of the proposed changes, although a Working
>>Group MAY issue a Call for Review. The modified Recommendation is
>>published according to the Team's requirements, including Publication
>>Rules [PUB31]."  That seems pretty simple as long as the change is fairly minor
>do it doesn't impact conformance.
>>>
>>> If it effects conformance, but isn't a new feature (so an
>>> implementation that
>>was conformant no longer is, or the reverse), it looks like it requires
>>implementations for what is changed and a 4 week AC review and Director
>>decision.  That seems pretty reasonable.  It doesn't look like anything
>>that shouldn't be necessary.  That's for an Edited Recommendation.
>>
>>Yes, that all seems rather reasonable to me. I've never pushed a PER
>>through, but I've never heard complaints that it was particularly painful either.
>>
>>> New features goes through the whole process to REC.
>>
>>That's not something we could possibly change since it has a direct impact on IP.
>>
>>> None of that seems bad - is it a problem to actually get through?
>>
>>I would tend to think that the biggest problem is making sure WDs are
>>reasonably up to date, far more than RECs.
>>
>>--
>>Robin Berjon - http://berjon.com/ - @robinberjon

Received on Friday, 10 February 2012 17:23:16 UTC