search

w3c / tr-pages

Work on the w3.org/TR index page (not specs themselves)
https://www.w3.org/TR/
26 stars 12 forks source link

ClearSpec 2021 #90

Open plehegar opened 3 years ago

plehegar commented 3 years ago

This is tracking issue for the project ClearSpec 2021.

The description of the project is at https://pad.w3.org/p/clearspec2021 (note that this will move into a .md next week to make the project public)

Comments on the overall project should be made here. Each separate project should have their own issue/milestone btw.

marcoscaceres commented 3 years ago

In the "use cases", it's unclear what:

Spec editors creating references to W3C documents

Means exactly, as it relates to a use case.

What does "misbehaviors" mean, in?:

a limited pubrules only intended to alert devrel on misbehaviors

About Project 2.

For "in document", ReSpec pulls data from "MDN data" already, and adds it to specs:

MDN

ReSpec also allows Editors to add caniuse data: caniuse

Do we want a global view/page/site of this data? The MDN data is generally authoritative, but caniuse is often not very accurate as it's very broad in what it means re: "feature support".

About:

  1. Consider extending this beyond browser implementations (ie going beyond MDN)

It might be good do consider clear differentiation of "data standards" VS "browser standards".

mnot commented 3 years ago

I'm excited to see the W3C paying attention to this topic. FYI, we're starting a similar discussion in the IETF here.

One of the lenses that I'd encourage people to look at this through is defensive branding -- i.e. "what do we allow the W3C identity to be associated with?" Confusion about the status and source of a specification isn't great for implementers, and it can be even more serious when the confused party is a competition regulator.

Some questions that might help:

  1. Should the W3C logo appear on documents that are not produced by W3C consensus (excepting things that are clearly not technical specifications, such as internal policy docs)
  2. Should the W3C's name appear on documents and efforts that are not consensus-based. In particular, why are they branded as 'W3C Community and business Groups'?
  3. Should W3C domain names (and other name spaces, such as on GitHub) be used for efforts that are not consensus-based? In particular, many people consider the domain name to indicate the primary publisher, and so it conveys a sense of authority, particularly to those who aren't familiar with the intricacies of the Process.

Overall, we should be more jealously guarding the branding of the Consortium. For things like CGs and BGs, a separate brand should be created, rather than diluting the clarity of the primary brand.

Note that disclaimers and statements in the negative (such as those you see in RFC and Internet-Draft boilerplate) do not work -- people get a document from a domain name, it has the associated logo at the top alongside a name, and they'll assume that it comes from that organisation no matter how strenuously you disclaim it in the boilerplate -- because most people skip the boilerplate by reflex.

marcoscaceres commented 3 years ago

Mark is right, but it's even worse... even without the logo, there is still confusion just in that something looks very W3C Spec... For reference, WICG drafts looks WAAAAAY too much like "proper" W3C specs - for example: https://wicg.github.io/cookie-store/

Even with the UNOFFICIAL plastered across the top.

Even a "CG-FINAL" spec looks too much like a real spec - for example: https://wicg.github.io/netinfo/?specStatus=CG-FINAL

We might need an entirely different color scheme or maybe even different fonts?

mnot commented 3 years ago

@plehegar has there been any progress on this work?

If we have proposals for the presentation of CG reports, etc., where should we post those issues?

marcoscaceres commented 3 years ago

@mnot https://github.com/w3c/clearspec