Re: Normative references to works in progress

TL;DR: This should be assessed case by case.


The long version:

The rational answer, and the one that drove W3C's old policy on how to  
link to unstable documents, is to asses each case and whether the status  
of the referenced document is accurately described in the document that  
makes a normative reference.

As far as I can tell this is still the sensible approach, so there is no  
general principle here beyond reflecting what we think the truth is. Thus  
far, I imagine everyone will agree.

However there is a fundamental disagreement on whether the core value is  
consensus in a multi-stakeholder model, or what three implementors decide  
is going to happen in their market-dominating software. So I expect in  
certain cases like the specific one at hand, the "right answer" will still  
be contested.

As far as I know, W3C's agreement with WHATWG is that while the parties  
can negotiate, at the end of the day W3C either accepts what WHATWG  
decides, or they can walk away from the arrangement. So unless the overall  
disagreement means the cost is higher than the benefit, the group should  
focus on determining what specific issues it wants to continue  
negotiating. In this case there might be a class of issues that are of the  
same kind, but I don't think there is a generally correct answer that can  
be applied across the board.

cheers

On Wed, 01 Sep 2021 03:21:15 +1000, Samuel Weiler <weiler@w3.org> wrote:

> I see that the WHAT WG HTML review draft has several normative  
> references to W3C documents that have not even reached CR, nevermind  
> full rec - docs that are in Editors' Draft status and thus might change  
> frequently, perhaps in breaking ways.
>  Should the parts of the HTML spec that depend on those non-CR documents  
> be marked in some way, e.g. by marking entire sections as non-normative  
> or subject to change?
>


-- 
Using Opera's mail client: http://www.opera.com/mail/

Received on Thursday, 2 September 2021 06:49:20 UTC