Should member submissions be removed from the Process? #648
Comments
|
As suggested by @michaelchampion, a member submission could be considered by a CG or WG no differently than any member contribution, with the exception of the requirements specified at Licensing Commitments in W3C Submissions. |
|
The licensing commitment is another anachronistic feature of Member Submissions ... .
Why should W3C perpetuate process features that don't require RF patent commitments on contributions? |
|
@michaelchampion do other parts of the process or the patent policy cover the situation where a contribution is the result of the work of many outside of W3C? Sometimes work starts within an industry group, who later realizes that the work has broader utility and would be best standardized through an organization like the W3C. It would be be good to have a way for such work to be provided as input into the W3C process. (edited with additional details) |
|
While member submissions aren't used terribly much, there are cases for which it isn't clear to me that there's a better replacement. Say you have a file format, which is a de-facto standard without published documentation. As something that is already broadly used, unless there's a will to make a next version with more features, there's not much point in spinning up a WG (or even a CG) to discuss how it ought to be, that's already settled by usage. But it could still be convenient for the company that made it to share the documentation for that format, along with information of what patent regime it is under. A member submission would seem to me to be a very appropriate way of sharing that. Members submissions aren't limited to this case, but they seem to me to be the only W3C process that's adequate for handling that sort of situation. For instance, see this note from https://www.w3.org/TR/css-ui-3/#cursor:
I'd say that it would be a nice thing for Microsoft to release the documentation for these formats as a member submission, and it isn't entirely clear to me that any other W3C path to publication that would be suitable. As for the patent regime, I'd surely prefer it to be RF, but what if, due to historical reasons on how the format was created, it cannot be? It might still be preferable to have documentation for the format and information about what patent regime it is under, than be left in the dark entirely. It could still be true that the cost of having and maintaining this system exceeds any benefit it may have. I am not sure I have that strong an opinion about that yet. |
Sure in 1995 or whenever the Process was defined, that might have been a valuable use case. But In a world where GitHub, Microosft/Google docs in the cloud, very low cost web hosting solutions, etc. etc. etc. exist why should W3C invest in sharing unvetted documents with the world? And if the answer is "submissions ARE vetted by the Director" then we have to address #639 and figure out how to get them authoritatively reviewed by someone else. Is it worth it -- to the people maintaining the Process, those who will have to do the reviews, and the community trying to understand the Process -- for the various corner cases that Submissions once addressed be supported in the future?
Right. I don't have a strong opinion, but the complexity cost seems to outweigh any benefit I've heard in this thread and #639. |
Possible reason: because these are about technologies upon which we have dependencies, whether we like it or not, and having them documented at a location which is guaranteed to be stable into the foreseeable future, along with a requirement for explicit documentation of patent licensing is preferable to having the same information hosted somewhere that might go 404 in 3 years, and for which there is no requirement of documenting the patent licensing. All in all, I find marginal utility in the availability of member submissions. But they don't seem terribly important either. |
|
Note that some sort of disclosure/publication process for specs that are relevant in the space is not completely unheard of. SMPTE has Registered Document Disclosure, for example; IETF has individual-track RFCs. Perhaps it would be good to encourage more use of this, so relevant vendor specs that materially impact our field are published this way? I'm not keen on spending a lot of time on either promoting this vehicle nor shutting it down, right now, though. |
|
The Independent Submission Stream for RFCs isn't part of the IETF -- it's a separate stream with an editor (who is selected by the IAB). That might get to the heart of why I'm wary of member submissions -- they have a smell of 'pay to publish' without any editorial oversight. In the past I've heard of them being used as an enticement to membership. Right now, s 10 puts the decision to acknowledge in the hands of the Team without any guidance (beyond making sure the request is 'complete'). It might be better to put that decision in other hands, or at least to provide more guidance about what the purpose of member submissions are, and therefore what the criteria for their acceptance is. |
|
I've not heard of any past instance where a Member Submission was blessed as a TR, instantly or otherwise, without being processed through at least a year — if not 2–4 or even more — of some IG/XG/CG/WG/whateverG(s) and being treated (mostly) the same as any other input document for that group's work. I have heard of a few (and been in such groups) where a Member Submission was published as a (no longer a thing?) First Public Working Draft of (one of) their anticipated final document(s) upon ratification by group membership, and bore some but far from complete resemblance to the final document(s) that went into CR/PR/TR. I can see some potential for membership enticement there, as I would expect it easier to get such a thing taken up as an input document when submitted by a member (so join!), not only but significantly because of the Patent Policy, which I think I remember takes effect upon such a Member Submission (differently than when a WD/FPWD/CR/PR/TR is produced entirely by the CG/XG/IG/WG/xG), than when submitted by a disconnected and apparently disinterested party (so why are you submitting to W3C?)... but I don't think this is any different than would be an available potion to any other member; this seems to me just "and don't forget, when you're a member, you can make a Member Submission to accelerate pre-WG (IG/XG/CG) work." |
|
There has been no member submission since 2020. I've changed my mind, and no longer think that this part of the Process justifies its continued existence based on some marginal benefit, even in a simplified form as proposed in #421. It seems to me that it is in everyone's interest to keep the Process short enough that it is approachable to new comers (Member companies and individuals). As we grow various parts of it in response to new needs, we should also be willing to retire obsolete parts, lest we force anyone diligent enough to try and learn the process to read pages worth or irrelevant rules and concepts. I think we should simply delete from the Process the entire Member submission section and subsections. This would not prevent the Team for offering a "we'll host your document if it seems valuable" service and document it whereever they want (/Guide?), but I don't think keeping it as a right of members with formal processes around it is relevant anymore. |
|
My sense is that the member submission process came out of a need, and even though that might be rarely used nowadays, I think the need probably does still exist. Before removing it, I think we need to understand what the expected model would be for a member to bring pre-existing work into W3C for use as the basis of a new Recommendation, and how issues like IPR would be handled. Perhaps there is a simpler alternative that does not require much, or any, spec text, but given that we don't have a proposal yet, I'd prefer to take no action, around the section as a whole. One area that might incur maintenance costs now is the one about rejection and appeals - I can imagine that refactoring that in terms of a Team Decision might simplify things a bit. I can't see a rationale for a whole separate Submission Appeals process now that we have Team Decision, assuming we can restrict visibility of such team decisions to the Team and the Submitter. |
Yes, but that predates CGs. Now, that's a much more common way to bring work to W3C.
Member submissions require you to disclose what your IPR is, but it doesn't require any favorable regime. Submitting something to a CG gives us more than that.
Something like that used to be my preferred approach. I no longer think it's worth the effort.
Note that I'm not necessarily saying we should discontinue the service. The Team could absolutely continue having this capability, and documenting how to submit things elsewhere (/Guide ?), which could be more focused on why you might want to do that, rather that the right and duties of various people. The Team provides IRC channels and all sorts of things without the being mandated by the Process. What removing it from the Process does is that it would no longer be a guaranteed right, letting us get rid of the even more unused appeals process, and no longer inflecting it on unsuspecting people. |
|
Perhaps I am opening a can of worms, but SMPTE clearly distinguishes the status of a member submission from work initiated and done at the consortium. They call theirs "Registered Disclosure" and say that at least one of the purposes is to provide the opportunity for members of the community to document practices that affect the industry and deserve public documentation, without any implication of endorsement by SMPTE. |
As suggested by @michaelchampion in #639 (comment)
@michaelchampion :
@dwsinger :
@michaelchampion :
@jeffjaffe :
@michaelchampion :
@mnot:
@michaelchampion:
@palemieux:
@michaelchampion :
@palemieux:
@michaelchampion:
The text was updated successfully, but these errors were encountered: