Re: marking obsolete TRs - ISSUE-2

On Fri, 02 Mar 2012 19:29:59 +0100, Philippe Le Hegaret <plh@w3.org> wrote:

> On Thu, 2012-02-09 at 19:33 +0000, Carr, Wayne wrote:
[in reply to]
>>> Instead of an ugly ignorable warning, you prepend "Obsolete" to the  
>>> spec's title, you make the background light grey, you change the top
>>> left W3C strip to grey, you bold the "Latest Version" link. The most
>>> practical would be to detect that in JS.
>>
>> Altering old publications that are obsolete to say obsolete is a good  
>> idea.  IETF does that, don't they?

Yes - although they have some automatic process that makes stuff hard to  
find if it *is* stable, so we should indeed be thinking about how to do it  
right. FWIW I like the suggestion above as what to do. Where it is a draft  
replaced by a new one, or simply abandoned, I think it is easy, but  
otherwise there is a question about who decides that matters to many  
people.

>> So only the latest version in TR would not say obsolete.  If a draft  
>> was abandoned before getting to REC, it could also be marked obsolete  
>> with a note that no further work was being done on that particular  
>> draft.

>> Along with this, if it is easy enough to publish a heartbeat draft,  
>> those could be done more frequently so not be so far out of touch with  
>> the editors draft.

IMHO (having done it in the past as a matter of course) it *is* easy. It  
takes some time, and there are editors who argue that an hour of their  
editing time is more precious than all the time all the rest of the  
working group spends. Personally, I think that only applies in totally  
dysfunctional working groups, and the worst cases I know of are only  
highly dysfunctional.

>> With anything without WG consensus marked as provisional.

Tracking things where there are issues in drafts should be pretty  
straightforward (for a "reasonable" level of quality).

>> It doesn't seem like any of this would require process doc changes (it  
>> doesn't conflict with the process).  Seems, the w3c staff could just  
>> agree to start doing it.

Actually, it just requires the working group to agree on a rhythm, and in  
particular the editor to act on it. While note free, I doubt taht it is a  
complex exercise for most cases.

There are two complications I see:

1. People creating specs for purely political purposes. There are numerous  
examples at W3C of people writing specs and attempting to introduce them  
as obsoleting something they personally dislike. Sometimes this includes  
good technical work, sometimes it's pretty much "Not Invented Here"  
allowed to run rampant :( Since the sensible updating of technology often  
comes from the "there are things we don't like" motive too, it is  
difficult to distinguish the two cases. But giving a handful of people the  
ability to claim something else is obsolete opens up a realtively  
straightforward gaming strategy, and we should assume it will be used, so  
think about when and how guard against it.

2. Where a working group allows the editor to determine what goes in a  
draft. In a model where the editor really edits, assuming that the working  
group generally contribute and decide efficiently, it is relatively  
unlikely that significant controversial issues are not lear in the spec.  
In the editor as author model that is becoming more common, it is easy for  
an editor to leave something out, and effectively leave out the issue  
markers as well. HTML5's issue 30 is a pretty clear example of this.  
Whatever the merits, the issue is invisible on reading the HTML5 spec, and  
has been kept that way for some years in what appears to be a clear  
editorial policy that has caused a lot of problems in the group,  
undermining confidence in the W3C.

> I would note that Charles added this issue into our tracker:
> [[
> ISSUE-2: TR documents which are obsolete or have been "parked" or
> abandoned... ]]
> http://www.w3.org/community/w3process/track/issues/2
>
> I'd like to encourage others to do so as well, both for the process and
> pubrules related items,

Yep. That way we can deal with them in some kind of rational manner.

cheers

Chaals

-- 
Charles 'chaals' McCathieNevile  Opera Software, Standards Group
     je parle français -- hablo español -- jeg kan litt norsk
http://my.opera.com/chaals       Try Opera: http://www.opera.com

Received on Monday, 5 March 2012 09:03:23 UTC