RE: w3process-ISSUE-126 (autoWDpublish): Automatic WD publishing tool may change the W3C center of gravity around WGs [Process Document]

Chris,
Well said, comments inline below:
Steve Zilles
From: Chris Wilson [mailto:cwilso@google.com]
Sent: Thursday, September 11, 2014 3:24 AM
To: Daniel Glazman
Cc: public-w3process
Subject: Re: w3process-ISSUE-126 (autoWDpublish): Automatic WD publishing tool may change the W3C center of gravity around WGs [Process Document]

There is no such "permanent right", as the WG (or the chair, or the team contact) could quickly remove an editor's assignment to that role (and their authorization to make such edits).

Sure, it is possible in this arrangement for an editor to publish changes that are not approved by the WG.  I would expect that editors to manage this appropriately; however, continuing with the system we have today has simply moved the bulk of specs I deal with to a system where the editor's draft is the place to go, because publishing /TR/ docs is needless paperwork.  (aka "Do I really care enough if my spec lives in /TR/ to jump through the same set of hoops regularly?")
[SZ] Based on the recent discussions in the CSS WG, editors do care if the TR document is up-to-date. That is because search tools (at least in the past) more often pointed searchers to the TR version of the document (more links to it) than the Editor’s drafts. That meant that new users and implementers were working with (possibly) out of date material and that has had unfortunate consequences. It is for this reason that Editors want to be able to update the TR page regularly.

Note that feel strongly that there is a responsibility that an Editor takes on, and as a spec goes to Last Call and beyond the consensus from the group should be stronger and stronger; but roadblocks we put in the way of improving the timeliness of /TR/ are not helpful.
[SZ] This was exactly the point some of the CSS Editors were making. There was a concern, however, that this means that it is difficult for a reader of TRs to tell whether a given version of the TR is an updated Editor’s Draft or an approved WG Draft. One suggestion, which may require a Process Change, was to have two levels of TR: WDs which would be approved by the WG as they are now and (for lack of a better name) Provisional WDs which would be more frequently updated by the Editor without requiring a formal WG action. (It was noted that we should be able to expect that Editors would be responsible and only make updates that would be unlikely to upset the WG without prior WG action.) It was suggested that these two kinds of TR would have different styling and status information, that Changes since last document would only apply to WDs, and that there would be two last version pointers in the header, one to the last Provisional WD and one to the last WD. This approach should work for both the audience (e.g. implementers) that needs up-to-date information and the audience that only want to review significant changes. The former uses both WDs and  provisional WDs and the latter would just use WDs. This seems to be an approach that satisfies both the concerns of Editors and the concern that Danial Glazman raised.

Regardless, the important part is a clear responsibility of the Editor as the spec approachs LC (and clear path of escalation if the Editor "goes rogue").  I believe this process has that.

On Thu, Sep 11, 2014 at 7:43 AM, Daniel Glazman <daniel.glazman@disruptive-innovations.com<mailto:daniel.glazman@disruptive-innovations.com>> wrote:
On 10/09/2014 21:45, Michael Champion (MS OPEN TECH) wrote:

>> If an editor 'goes rogue' and starts publishing what appear to be WG WDs without the WG agreement, I rather think their life as an editor may be short,
>
> +1.  I think Chairs have the tools they need to handle this situation. Plus with the one-click publish tool, an editor can't hold the chair hostage by saying "the publication process is so complex, it will take you forever to find a replacement if you fire me" :-)

I don't think these two answers quite address the concern, sorry: the
concern is to see a given WG allow one editor a permanent right to
publish his Editor's Draft as a WD. That's a de facto 'living WD' and
the WG is not in charge of the publication any more. All of that is
technically possible and I think it _will_ happen in the future if that
posssibility is open.

Sorry, but I am very, very far from excited by the potential
side-effect of that tool...

</Daniel>