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

Reference to W3C Group in standards/types #102

Open csarven opened 2 years ago

csarven commented 2 years ago

Editor's draft states:

An Editor's draft is a document produced by a W3C Group.

To be clear, should "W3C Group" link to https://www.w3.org/2021/Process-20211102/#w3c-group or something else?

If so, are Community and Business Groups discouraged or prohibited from publishing documents with "W3C Editor's Draft" (in content / visuals) status?

Should CGs and BGs be recommended to use "Editor's Draft" without the "W3C" (content / visuals)? I understand that "Unofficial Draft" is arguably the closest alternative, but as far as I can tell, documents published under CGs use both ED and UD (and sometimes even interchangeably.)

If the text reference to "W3C Group" is intended to be any W3C group, I suggest clarifying this.

plehegar commented 1 year ago

The text was meant to say "any W3C Group" but I also believe there is value in reserving "Editor's Draft" to Groups that are only defined in the W3C Process, per 3. Groups and Participation.

@marcoscaceres , got an opinion?

csarven commented 7 months ago

@plehegar @frivoal @fantasai @koalie , perhaps the relevant parts of this issue should also be followed up in https://github.com/w3c/w3process (since it touches both standards/types and Process)?

koalie commented 7 months ago

If so, are Community and Business Groups discouraged or prohibited from publishing documents with "W3C Editor's Draft" (in content / visuals) status?

Should CGs and BGs be recommended to use "Editor's Draft" without the "W3C" (content / visuals)? I understand that "Unofficial Draft" is arguably the closest alternative, but as far as I can tell, documents published under CGs use both ED and UD (and sometimes even interchangeably.)

CGs/BGs are expected to follow the reports requirements.

koalie commented 7 months ago

If the text reference to "W3C Group" is intended to be any W3C group, I suggest clarifying this.

The document already lists in the summary table (section 1) all the types of documents and whether their work is endorsed by W3C or not; and the next section (2.1) is specific to CGs and BGs. I just amended the content to add links to the CG/BG Process, and the Reports requirements document (which in turns refers to which stylesheets to use.)

I'm open to suggestions to make this even clearer if you have ideas.

plehegar commented 7 months ago

so, if I understand this right, "Editor's Drafts" are reserved for Groups under the W3C Process (ie, not for CG/BG). Correct?+

marcoscaceres commented 7 months ago

Sorry for the late reply, I don't think there CG/BGs should be using "Editor's drafts" (unless there are exceptions) - we have CG-Draft and BG-Draft status for these, right?

frivoal commented 6 months ago

The only thing the Process has to say about EDs is:

Working Groups and Interest Groups may make available Editor’s drafts. Editor’s drafts (ED) have no official standing whatsoever, and do not necessarily imply consensus of a Working Group or Interest Group, nor are their contents endorsed in any way by W3C.

I guess it can be understood in two different ways:

I guess my view has been closer to the second: when you look at an ED, you're looking at one person's take (the Editor's) on the work, which is in the process of being prepared, hopefully soon to be adopted by an official group according to some kind of rules. Once it does, it will have some kind of status, denoted by a particular name, but until then, it's just some person's draft. I guess the difference I make between an ED and an Unofficial Draft is that for an ED, there needs to be some kind of group who has deemed the person to be an Editor of a thing. So the document may have no status, but the person making it does, to some degree. (But that only shifts the question: what kind of group is official enough to be able to deem people to be Editors?)

However, given that for many groups, official drafts tend to lag severely behind, EDs have become recognized as the place you go to find the most up to date version of some group's work. I think that's what WDs were meant to be, but here we are.

Under that view, the question becomes: do we try and put WDs back into the role of being the most up to date version of a group's work, in which case EDs can go back at being one (appointed) person's take, and can be used in any context, or do we accept that EDs are semi official Working Draft Draft, and should only be created by groups that can made Working Drafts (i.e. WGs)?

I'm more of the former, but I do recognize that's a bit of an uphill battle.

csarven commented 6 months ago

Restricting the use of only DRAFT/FINAL reports for CG/BGs goes against the idea of nurturing incubation spaces and is not practical. And, evidently, CG/BGs do use ED/UD document types out there as part of their work mode, prior to committing to publishing DRAFT/FINAL reports (which carry partial patent protection). ED/UDs enable CG/BG participants to advance work smoothly without unnecessary procedural or legal hindrances, and without interfering with the idea that they are not endorsed by W3C or part of the standards track.

If ED/UDs are intended to be reserved for https://www.w3.org/2021/Process-20211102/#w3c-group as I've mentioned earlier, so be it, but this ought to be even more clear in https://www.w3.org/standards/types (and elsewhere).

I still maintain that CG/BG participants, especially editors, can use an equivalent to ED/UDs. Expecting them or the Groups to solely publish DRAFT/FINAL reports complicates matters and involves the entire Group, whereas displaying ED/UDs of work in progress DRAFT/FINAL would be a simpler approach. This wouldn't necessarily be an ideal solution, which is why the simplest and most intuitive would be to clarify that ED/UDs can be used by any Group.

frivoal commented 6 months ago

prior to committing to publishing DRAFT[…] reports (which carry partial patent protection)

Are you sure about this? I believe that contributing material to a CG activates those patent provisions, regardless of how things get published. The words "Draft report" do not appear in the Community Contributor License Agreement. It does speak of contributions to the Specification, but with no condition on its publication status.

csarven commented 6 months ago

My reference was to the overview of documents under https://www.w3.org/standards/types/#x1-summary for row "CG/BG Report", column "Patent Policy" with value "Partial". Happy to be corrected on the wording I used, and that whether "partial"

Right that Community CLA applies to items that are acknowledged / maintained by the CG/BG. Note however the row "ED", column "Patent Policy" with value "No".

ED/UD(-like) document types can enable CG/BGs to show/signal work in progress towards DRAFT/FINAL reports. There is a specific process to publishing DRAFT/FINAL reports and they carry more weight than ED/UDs. No significant "decision" is required for ED/UDs beyond those afforded to editors/authors as part of their role in any type of Group.

frivoal commented 6 months ago

(I am not a lawyer, but) my understanding is that Patent coverage works very differently in WGs and in CGs. In WGs, it is based on the specification hitting certain milestones, at which point everybody must provide a license to everything in the spec (unless they exert their right to exclude). ED is not such a milestone, so content that is only in an ED does not get coverage by the Patent Policy applicable to working groups. So in that sense, the summary you pointed to is right.

However, CGs do not work based on milestones, they work based on what you contribute yourself. When you contribute something towards a specification, regardless of how you make that contribution (email, github PR, write an ED on your own…), you must provide a license to what you contributed. (There's also the Community Final Specification Agreement, but that's entirely opt-in, so it's not relevant here.)

Now, if EDs are things that can only be made by Working Groups, then the claim that they have no patent coverage is true. But that's because of how WGs work, not because of what EDs are. If EDs can be used in CGs, then to the extent they contain contributions to the work of the CG, then they do have some patent coverage.

marcoscaceres commented 5 months ago

Unfortunately, EDs drafts are used as canonical things in the community, so they have more official standing than we would like. For example, HTML exclusively (and by policy) only points to EDs and shuns TR documents. Thus, to a lot of the community, they represent more than some editor's opinion.

Further, some working groups (e.g., WebApps, some of Web App Sec) only land things in the Editor's draft with consensus and implementation commitment... (which then get automatically published as Working Drafts on TR).

That's not to say that some other groups don't treat EDs with little care and as low value drafts... but let's please remember that EDs are treated as canonical by a lot of folks.

csarven commented 1 month ago

@koalie , thanks for the changes that you've mentioned in https://github.com/w3c/tr-pages/issues/102#issuecomment-1794692196 . I do think that helps and I acknowledge the context in which "W3C Group" is used in that it falls outside of "Pre-standardization, proposals, notes".

My initial concern was that all things considered, the term "W3C Group" being limited to only some types of groups at W3C is not immediately clear anywhere. I think anyone could easily assume it could mean any type of group at W3C - I know I have in the past. And people do use "W3C Group" liberally, where without context, it can be misleading. This is what I was hoping to clarify. I don't know where best to make the distinction - to spell it out. Don't have a strong opinion on it either. Perhaps a top level sentence at https://www.w3.org/standards/types/ or in its Summary?

I've been contemplating the role and usage of ED/UD in comparison to the role played by CG-DRAFT. One o the things that caught my attention is that the default stylesheet for CG-DRAFT applies an "unofficial" watermark, which is removed when the no-watermark class is applied to the <body> element. Please correct me if I'm mistaken, but I assume that the transition from watermark to no-watermark occurs when the CG requests to publish a CG-DRAFT and once it is approved by the W3C Team Community. That aside, it seems to me although this might be obvious to some but not detailed anywhere as far as I can tell in that CGs can use CG-DRAFT in unofficial state where they can use the document along the lines of an ED/UD (without of course saying anything in copyright or SotD that would conflict with W3C CG Process of course). I believe that simple clarification would not require any significant changes to process or rights.

Does that sound about accurate and reasonable? If so, could this kind of a clarification between unofficial CG-DRAFT and the official CG-DRAFT, i.e., approved and published be added to https://www.w3.org/standards/types/#x2-1-w3c-community-group-report-or-w3c-business-group-report ?

As I've mentioned earlier, I do believe that CG/BGs can use some assistance on what's an official report (whether DRAFT or FINAL) from the unofficial for their day to day WIP stuff - which I think is analogous to ED/UD pragmatically speaking.

csarven commented 1 month ago

Correction to my "no-watermark" above:

after re-re-revisiting, comparing tool outputs, checking CG reports, reviewing styles,... my current understanding is that CG-DRAFT only comes with a watermark and the no-watermark is an isolated case in https://w3c.github.io/rdf-star/cg-spec/2021-12-17.html (besides the logo being stuck on CG-DRAFT while content saying "Final"). I do not know if the use of no-watermark was allowed by W3C Team, or an error / oversight somewhere.

koalie commented 1 month ago

@csarven I just spent a half-hour re-reading this issue's content and I think no further edit to https://www.w3.org/standards/types/ is necessary (but if you think differently, I welcome proposed edits).

For your questions about the CG/BG reports and in which conditions "no-watermark" is allowed, I call on @ianbjacobs to clarify here and possibly in the CG site documentation too.

csarven commented 1 month ago

@koalie , I'm not sure if this is the correct repo to update the standards/types page but I've created https://github.com/w3c/tr-pages/pull/108 for your consideration.