RE: Transition to a revised Technical Report Development Process [W3Process-ISSUE-39, W3Process-ACTION-10, proposal]

Elika,
I like your desire for simplicity, but as usual the devil is in the details, see below.

-----Original Message-----
From: fantasai [mailto:fantasai.lists@inkedblade.net] 
Sent: Wednesday, October 30, 2013 2:49 PM
To: Ralph Swick
Cc: Advisory Board; W3C Process Community Group
Subject: Re: Transition to a revised Technical Report Development Process [W3Process-ISSUE-39, W3Process-ACTION-10, proposal]

On 10/14/2013 05:55 AM, Ralph Swick wrote:
> Re: ISSUE-39: Managing the transition to a new TR cycle
>
> Should the W3C Advisory Committee approve a new Technical Report 
> Development Process the Director will need to state the manner and 
> schedule for deployment of the revised Process.
>
> As a stake in the ground for discussion, I propose the following:

Here is my alternative proposal...

As of the official transition date chosen by the AB/AC/W3M/whoever:

   1. Any spec in WD, CR, or REC is automatically transitioned into
      the new rules for WD/CR/REC, without republication or any WG
      action.

SZ: Certainly any REC can move to the new process. In principle, any WD also can move, but there clearer requirements (SHOULDs) that are intended to keep reviewers of the WDs more informed about what had changed so moving the the new process even for a WD may involve some extra work. This is a good thing, but is more work for publishing a next draft. CR, however, is not easy to move to the new process because CR and PR are being parallelized in the new process and current CR specs have not had the done. The current suggestion is to let anything in CR finish under the existing process, hopefully without return to LC, etc.

   2. Any spec in a transitional phase (LC/PR/PER) follows the current
      process until it gets to its next stable phase (WD/CR/REC),
      and from then on follows the new rules for WD/CR/REC.

SZ: I would have to think more about the LC case, but since the next phase for PR and PER is REC, hopefully and, by experience, actually, this may be reasonable. The main caveat I would have would be that if for some reason a PR or PER is sent back to LC, then it should complete under the old process because there is no need to have the editor (and WG) learn a new process just to make the small updates that typically require an extra LC.

Are there any reasons why this isn't complicated enough?

The reasons are suggested above. To me the main issue is when is the conversion to the new process to happen. There is a contingent that believes that a WG that is nearing completion on one or two documents ought to be allowed to complete that work under the process they currently understand. It is also recognized that there are groups, such as CSS, HTML5 and WebApps, that have documents in a number of states and will just have to byte the bullet sometime. For these groups, I think you suggestions (with the caveats I raised) make sense.

Steve Z

~fantasai

Received on Thursday, 31 October 2013 04:32:28 UTC