[decentralized] Propose that the name be changed to avoid overlap with "dWeb" #281
Comments
|
Strong +1 that "Decentralized Web" means something already, and it does not appear to be what this group is chartering. |
|
Some other confusingly used terms in the draft questionnaire from which the group proposes to start work:
Avoiding name collisions would help to improve shared understanding. |
|
We do need to avoid namespace collision primarily within the W3C. Other candidate options considered were “interoperable”, "impartial", and “stable”. “Interoperable” was dismissed as it is widely used already within the W3C. "Impartial" dismissed as it lacked clarity. “Stable” dismissed as it poorly reflects only one aspect of the proposed groups purposes. The following was considered during the drafting process.
Given the feedback and the considerations I propose the name “Decentralized Interest Group”. The word “web” is not needed in the context of the W3C, and it clearly separates the group from the externally named from DWeb. DIG would be the acronymn, or W3C DIG when used externally. |
|
@wseltzer regarding the draft questionnaire. Great suggestions. There is a pull request on those documents. I'll seek to include the resolution of those namining collisions there. |
|
@jwrosewell unfortunately I think that using 'decentralized' in the name in any way is still going to be confusing. Merely dropping ’web’ does not make things more clear for me, at least. My understanding is that this proposed group would be looking at ways to include considerations for systems that interact with the web, and possibly devolve some decision-making processes away from browser vendors. I think devolution has some baggage, but perhaps it would make sense to have a “Web Ecosystem Interest Group” or ”Web Interconnections Interest Group” |
|
What about "Stakeholder Representation" or "Web Interconnection" as @astearns suggests above? We're still hearing from people who find it hard to discern the group's goals from the name or even from the charter, so while naming things is one of the known hard problems of computer science I think it's worth the effort to clarify before the charter goes for review and gets objections on that basis. |
|
I'm not sure what "Stakeholder Representation" is intending to mean, but it seems to be more making a comment about who is allowed to participate in other standards than it is an independent group. (And it's not a positive implication, nor one that I think is true.) I would say that the charter of this group seems more like "Web Ecosystem" than anything. I still don't understand the goal of having it as an interest group, as the horizontal review aspect - to the extent that this function is not simply part of building any web standard to begin with - seems to fall under the auspices of the TAG today. |
|
The name of this group should readily and succinctly describe the group mission. We concluded the group name must be unique within the W3C. We have observed complexity associated with avoiding naming conflicts outside the W3C. Therefore, prefixing the group name, or any group name, with W3C avoids that conflict. This approach is commonly used. Examples include W3C Technical Architecture Group or W3C Automotive Working Group. Each of these names would carry other connotations outside the W3C should the W3C prefix be removed. As the World Wide Web Consortium (W3C) already includes the word web, and is itself well known within the field of technology standards, dropping the word “web” from the group name is unlikely to confuse anyone. The mission of the group is to improve recommendations to further the W3C’s One Web purpose by identifying, balancing and mitigating the unintended impacts of proposals early in the process, encouraging uniform change, and promoting OpenStand principles. Many proposals overlook or discover late in the process the unintended consequences that impact stakeholder groups that are underrepresented within the W3C. Thus we aim to assist specification authors in identifying, balancing and mitigating these consequences, specifically around the changes that would centralize control rather than support the decentralized, One Web. Explicit support for the charter as drafted and named includes SafeCast, which protects children and other vulnerable people online, the German Association for the Digital Economy (BVDW) [members], European Publishers Council [members] and many W3C members and other interested stakeholders. It is for all the above reasons the word "decentralized" is favored and the new group name will become the Decentralization Interest Group or DIG. Pull request 293 includes this change. A combined list of BVDW and EPC members follows. |
|
@jwrosewell I do not think you have addressed the issue raised here, or been responsive to the alternative suggestions provided. A separate suggestion on issue markupYou can use markup to nest supporting content, as I have done here. I don’t know whether you’re asserting that all members of this list require the name you prefer, but the long list isn’t particularly compelling without showing their reasons. I would prefer you delete the list or move it to a `details` element to make the relevant parts of issue easier to read. |
|
@astearns thanks for the comment on collapsable markup. Changed. Always learning. In relation to your comment on addressing issues I have covered this in more detail in another issue. In relation to your comment here.
Perhaps I misunderstood this sentance. Certainly in relation to the second part of the sentance that is not the case. Browser vendors remain free to do as they wish. I don't feel this charter, or any charter, needs to state this explicitly. The W3C Process applies to everything. Could you elaborate?
The purpose of the group is to identify unintended consequences with a specific emphasis on ensuring the open web is decentralized and furthers W3C One Web mission. As such I, and other proposers, feel that "decentralized" remains the most appropriate group title. |
|
Do you think your proposed group matches the scope and intent of the groups mentioned in the first comment that already use the term ‘decentralized’? So far, I don’t see how the new group relates to the existing groups. I don’t think it’s the right move to take over the term in the W3C context when it means something different outside (but very relevant to) the W3C. |
The groups mentioned at the start of this issue appear to have concerns that the proposed group could assist with and might represent a subset of considerations. The new group is concerned with all aspects of decentralization and unintended consequences. This is the only relationship I can identify between the groups. As the none of the existing groups referred to operate within the W3C, there is clear precedence for name overlap outside the W3C which is dealt with via the W3C prefix externally, and that there is overwhelming support for the group charter as drafted and named I would prefer not to delay progressing the formation of the group due to this issue. As such I think the next step for this issue is with W3C Team to advise if the precedent prevails or if an exception is needed in this instance. |
|
@jwrosewell I don't understand your reference to precedent. Multiple people, including the W3C Director, have said they find use of the term "decentralized" confusing. Team would prefer to address that confusion before proposing a group for chartering. |
|
Perhaps "Decentralization" rather than "Decentralized"? In relation to precedent there are plenty of examples of naming overlap outside the W3C. "Automotive", "Technical Architecture" being two. |
|
The draft charter says the mission is "identifying, balancing and mitigating the unintended impacts of proposals". Shouldn't Unintended Impacts be the name of the proposed group? Problem is, it's VERY hard anticipate unintended consequences. After all, the web itself evolved out of an academic community that used URL/HTTP/HTML to share new information and discuss theories to explain it to benefit humanity; 30 years later the web is overwhelmed with disinformation and propagation of conspiracy theories to promote various bad actors' self interest. If only we had somehow anticipated that that same platform that allows loving couples to find each other would allow haters to find each other and gang up on their victims, and somehow have tilted the platform to favor the lovers over haters, truthseekers over liars, and traders over the thieves! Alas, we humans don't really have reliable ways to anticipate unintended consequences. I think it would be great to setup some sort of group to explore what we could do, probably a CG, since we'd want broad participation from multidisciplinary experts who may not be W3C members and can't justify the high cost of joining. But that's not what this charter proposes. Throughout the charter is an implicit assertion that "decentralization" is a core value of the web, so we need to anticipate threats to that core value. If the mission of the proposed IG is really "identify proposals early in the process that are incompatible with a decentralized web", just say it. But it's not clear to me how to prevent standards whose implementations aren't subject to the economies of scale, exponential increase in their value as the number of users rises, and the other real-world "laws" of web economics. Bottom line for me (personally, not speaking for any W3C member I am or ever was affiliated with): This is a FASCINATING topic, and worth much research, analysis, and discussion... but it would be very premature to form an IG to act as a gatekeeper to apply its members beliefs and interests in hopes of mitigating the problem of web centralization. |
|
@michaelchampion I agree that it is impossible to identify every unintended consequence that will risk centralizing the web. That does not mean we should not encourage new proposals to address this risk. Similarly, the Privacy INterest Group (PING) references a none exhaustive list of harms captured across a wide range of documents. The draft documents that preceded the proposed establishment of this group do not describe right or wrong. Instead they provide a framework to help identify issues that are often overlooked. The group will strive to include a diverse set of members and skills to better represent all web stakeholders. This group will not preference one set of beliefs over another, but merely ask proposers of new standards to address the interests of underrepresented stakeholders. If a group were to preference specific beliefs that undermine the W3C One Web mission it would be in breach of the W3C membership agreement. |
|
I understand but disagree that this proposed charter would be an effective way to ensure that the interests of under-represented stakeholders are considered in W3C. My suggestions for improving the charter are:
|
|
To add to Mike's comments (which are 💯), I'd point out that we had a similar objection at the last rechartering of the PING - that we believe that the PING should be responsible for some detailed threat modeling, to give those developing specs some idea of what exactly the PING would be looking for (aka "teach spec authors to fish, rather than throwing the occasional fish at them"). I'm not sure what you're trying to imply by "If a group were to preference specific beliefs that undermine the W3C One Web mission it would be in breach of the W3C membership agreement." Perhaps you should clarify. (The Membership Agreement https://www.w3.org/2009/12/Member-Agreement doesn't mention the W3C's mission at all, nor One Web.) |
|
The proposers of the charter worked on the documents in the Improving Web Advertising Business Group (IWABG) where it was felt they related to more than advertising and as such should be rehomed. The participants of that group met and agreed to pursue an Interest Group charter to progress these goals and create documents with official standing inside the W3C to aid others. The PING charter and the feedback referenced by @cwilso (which are member confidential) was consulted when drafting the charter. It is for precisely the reason @cwilso highlights the group are producing documents, and providing expertise on request, but not involved in horizontal review as the PING does. DIG is all about “teaching others to fish”. Where matters are presented for AC vote the group do intend to provide a short summary to aid in highlighting unintended consequences for time poor AC members. This may help increase the number of members participating. The W3C Membership Agreement describes a purpose that could be summarized as the One Web mission and is summarized as such in the Mission statement. The full text from the membership agreement reads as follows.
I would hope we can all agree that the One Web mission nicely summarises the Membership Agreement. If not we should look to address this in the Process 2021 work. |
|
I don't think One Web has anything to do with the Membership Agreement; nor is it in the Process document. The W3C's strategy and vision are determined by the Director and the Team; with advisement from the TAG and AB, of course. The Process defines the mechanics by which specifications are developed, how WGs work, etc; the Membership Agreement defines the mechanics by which... well, by which the W3C takes Members' money and enables them to participate. (Note that if you're referring to Member confidentiality because you think I'm violating it 1) I was specifically referring to Google's high-level feedback on that charter, which I authored, and thus can "declassify", and 2) that object actually became a matter of public record anyway, as formal objections that are not resolved do. If your goal is to write documents to teach others to fish for your particular kind of fish, by all means have at it; but a better type of group for doing that would be a Community Group. If you intend to comment on every AC vote to "aid time poor AC members", you can do that as an AC forum member. |
|
I do not think @cwilso violated member confidentiality. I added the point in brackets for the benefit of those that are not W3C members but are following this issue. If the membership agreement does not reflect the mission of the W3C then that is an issue for the Process CG. I have raised that as an issue titled "Align membership agreement to W3C mission" in Process CG where that can be debated. It does not appear material to chartering this group or the issue of name. I’m finding inconsistencies in the responses to this group draft charter and the PING charter that has already been granted. As this topic relates to the name of the group rather than the work product, skills of participants and activities of the group I’ll continue the discussion on the issue related to clarifying intent. We can return to the subject of group name once the other matters are sufficiently progressed. Are others comfortable with this direction? |
To be pedantic :-) it’s actually an issue for the W3C management team, the Host universities, and the AB who are figuring out how to structure a new nonprofit corporation that will take over the role the Hosts have played. As I understand it, the Process CG is mostly about implementing the guidance of the AB (which formally owns the Process document) in language the community can understand and live with.
The PING charter was somewhat controversial. The pushback you’re getting seems to be against using PING as a template/precedent for other groups that hope to do “horizontal” review of specs.
I agree that #291 is a better issue in which to discuss the mission as opposed to the name of this WG. |
|
Yes, let's move the conversation to #291 re: mission. On the particular topic of the PING charter: the PING group had already taken up the mantle of reviewing specifications - and that is their primary role. They drive this - they proactively look at new incubations and specs, and also respond to requests for review (but generally, they get to it first, in my experience). The PING rechartering last year was particularly contentious because Google made a point of asking that the group actually be chartered to document the guidance by which they would review specs - i.e., that they detail the kinds of threats to privacy that they would be looking for, so spec authors could "learn to fish," as it were. However, I don't think anyone involved in PING would disagree that that documentation is a secondary function of PING, second to their role of regularly proactively reviewing specs. That doesn't seem to be how you're thinking of this group, but let's take that discussion to #291. |
|
Late to the discussion, but as a member of decentralized web community (disclosure: working on @ipfs project) I'd like to +1 to the concerns related to the "decentralized" name. Really suggest picking something that can't be confused with existing dweb movement. The "decentralized/distributed web" (DWeb) movement was build on values of user agency, data control, privacy and trustless security models, which today are often in direct conflict with actions, priorities and incentives (not words) of adtech and enterprise. Reappropriation of "decentralized" term here sounds really disingenuous, especially in the context of discussion about intended activities happening in #291. Even if the name choice was made in a good faith, as years go by this will cause not only confusion, but bring serious skepticism around the true agenda of the group, losing good will which would not be the case if a more honest name was selected. Some materials in addition to ones from #281 (comment):
|
Hi all,
Just wanted to raise the issue that the title of this group is very misleading since there is a well established DWeb movement that uses this term explicitly.
See:
And the search results at: https://www.google.com/search?q=decentralized+web&sourceid=opera&ie=UTF-8&oe=UTF-8 from where those links came.
Naming things is hard as we all know, and I'm not coming in with an alternative name right now, but as someone with an interest in both this group and the "dWeb" movement I find the overlap pretty confusing. It might cause issues if, as I assume it will eventually do so--if it hasn't already, the W3C opens a group to talk about the dWeb version of this name.
It seems like a namespace collision worth avoiding!
The text was updated successfully, but these errors were encountered: