Re: Normative references to works in progress

Additionally, I think we've not been as good as we'd like about keeping the copy of the document on TR as up to date as it should be. If the WHATWG choses to link to a document that is pre-CR, it's their choice (and we can question in along the lines of what Chaals mentioned), but so be it. But when they chose to link to an editor's draft for a document that has a copy on TR, it reflects more on our ability to keep the official copy up to date than on our differences of policy about what gives a draft its standing.

There are probably a number of documents for which the TR copy is up to date and for which they still link to the editor's draft, but that's to no small degree due to the fact that we have indeed been bad at keeping things updated. I think we're getting better (partly through better process, through better tooling, and better habits), and as individual documents do become well-maintained on TR, it is reasonable to ask the WHATWG to start pointing to the proper place. But I would expect some inertia there as well.

—Florian

> On Sep 2, 2021, at 15:48, Charles 'chaals' (McCathie) Nevile <chaals@yandex.ru> wrote:
> 
> 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 Sunday, 5 September 2021 12:15:52 UTC