Re: Consensus on ISSUE-73 (was Re: Universal Property)

On 18 Jan 2008, at 17:04, Michael Schneider wrote:

> Please for my education:
>
>   <http://www.w3.org/2005/10/Process-20051014/ 
> policies.html#FormalObjection>
[snip]
> I do not really understand what exactly the director (TimBL ?)  
> will / can /
> should do in such a case. Can anyone answer, or better, give an  
> example of a
> previous formal objection and how it was handled?

A formal objection merely means that the content of the objection  
must be reviewed by the director before changing the status of the  
document. The chairs and staff meet with the Director and discuss the  
objections, their technical content, and the other factors (such as  
group resources, time to market, etc.) that led the chairs to proceed  
over dissent. The Director can refuse to advance a document over a  
formal objection, in which case, the group might have to go back a  
level (e.g., from CR to LC).

OWL 1.0 was approved over formal objections (including, IIRC, by Jim  
oh...and by Jeremy!):
	http://www.w3.org/2001/sw/WebOnt/webont-issues.html#I5.6-daml- 
imports-as-magic-syntax (Jim)
	http://www.w3.org/2001/sw/WebOnt/webont-issues.html#I5.26- 
OWLDLSyntax (Jeremy)

I won't editorialize specifically on the content of these objections  
since I think it's pretty clear why they didn't have any effect and  
that it was right (and easy) to override them. Tim's pretty sensible  
that way.

I know that other specs have made it up to PR and then had to go back  
to LC. Sometimes that is done to head off formal objections, though  
in those cases the objections seemed addressable.

WSDL at various points had up to 3 formal objections, two  
diametrically opposed with folks like IBM, Sun, Microsoft, etc. lined  
up on opposing sides ;)

So, I wouldn't get too worked up about formal objections. We are  
obliged to try to seek consensus, with a preference for unanimity.  
But, for example, in this case, Manchester and several other groups  
would clearly formally object to making owl:Thing necessarily  
infinite while pointing to the complete lack of use cases, the  
implementation burden, the divergence from standard formalism, the  
breaking of services, the backwards incompatibility, the existing  
workarounds, etc. etc. I know which overriding I would prefer to  
defend were I chair :)

Cheers,
Bijan.

Received on Friday, 18 January 2008 17:30:53 UTC