W3C Community group

[Gozala]

At the dweb summit browser panel we talked about idea of starting w3c interest group with a focus of figuring out what APIs will need to be added to the web platform to allow dweb applications with no server requirements.

Here you can find all the existing w3c community groups:
https://www.w3.org/community/groups/

Looking through existing groups it appears to me that Decentralized Communications group has a similar goals and might be worth joining that group instead of spawning a new one.

I can also see how that group has similar but different goals so spawning a separate group might be a worthwhile.

Before we do start a group I'd like everyone to provide your input on:

• What is your vision of the goals.
• What is success in your point of view.
• Is d in dweb stands for decentralized or distributed, do we need diff term ?

[Gozala]

I'll start by answering those questions myself:

What is your vision of the goals

My goal is to extend web platform in a way that would enable ecosystem of applications that do not require web-server. Instead web applications would have an ability to establish p2p connections for data exchange and replication.

What is success in your point of view

Community defined set of high level APIs that could enable web applications that are capable of

• Inviting participants on define participation terms.
• Allow participants to collaborate both in realtime & offline.
• Allow data exchange without caring about transport layer.

Is d in dweb stands for decentralized or distributed, do we need diff term ?

I personally have being avoiding defining d because different people put different meaning. Often times decentralized is used for blockchains even though it's logically centralized. What I care about is a web where participants can collaborate regardless of network topology or lack of network thereof. In that sense p2p captures it better although sadly has bad connotation.

[lidel]

What is your vision of the goals

My goal is to identify and remove single points of failure from web platform by:

• making web browser usable in contexts with limited connectivity to the internet by taking advantage of local peers and network topology available at hand
• removing the need for centralization and trust (where possible) by means of things like content-addressing

What is success in your point of view

• web browsers support p2p communication (discovery, transports) out of the box
• new protocols coexist/compete on equal grounds thanks to generic APIs provided by web browsers
• decentralized and/or trust-less alternatives to the current HTTP+DNS+CA/PKI stack exist

Is d in dweb stands for decentralized or distributed, do we need diff term ?

Terms are loosely defined: dweb looks good, but "p2pweb" is more accurate description of the endgame I work towards. First step is to decentralize, enable things like federation. Then make it robust and go distributed by enabling real p2p communication.

Write "dweb", think "p2pweb".

[Gozala]

Hi Folks,

We've drafted a charter for the community group, looking forward to your feedback
https://via.hypothes.is/https://gozala.hashbase.io/posts/Peer-to-Peer%20Web%20Community%20group/

P.S.: Posting hypothesis link to allow inline comments / feedback. Version without third parties can be accessed here https://gozala.hashbase.io/posts/Peer-to-Peer%20Web%20Community%20group/

[lgrahl]

I've added a couple of comments. But I want to emphasize one in particular: I would rather do this work as part of the WHATWG if possible. The charter states...

The group will foster community agency: creating conditions for civil discourse, human dignity, and individual expression independent of corporate and legal structures. [...]

As proposals gain support and become more stable and mature, they will be considered for migration to a W3C Working Group [...]

The organisation structure of the W3C pretty much contradicts these principles as being a member in the W3C usually requires paid membership.

[Gozala]

@lgrahl Thanks for the feedback. I've made update to incorporate your feedback, please let me know if that resolves your concerns or if you there's anything else.

[lgrahl]

LGTM. I guess the question that remains now is under which umbrella organisation we want to do this.

[tantek]

https://github.com/lgrahl, by “which umbrella organisation”, which organizations do you think are setup / intended to handle incubation efforts that are not yet ready for standardization? Or do you think there is already something that has been incubated that is ready for standardization? If so, could you link to what in particular and where it was incubated?

While I am a fan of doing standardization work @WHATWG (and @W3C and @IETF, depending on which is most appropriate accordingly), I will point out that WHATWG themselves note that they “anticipate features will be “incubated” outside of WHATWG standards” per https://whatwg.org/working-mode#new-proposals

You also referenced W3C’s paid membership model. Note that W3C Community Groups (which this is a proposal for) are open to all (anyone may participate) and without fee (no membership is required) per https://www.w3.org/community/about/

The existing W3C CG WICG https://www.w3.org/community/wicg/ has a fairly good track record for incubating proposals in their GitHub repos which then either graduate to a working group (some have gone to W3C, others to WHATWG), or incubation stalls/fails as a result of discovering an area or proposal is unworkable or lacks market interest etc.

Are there other incubation-friendly/centric organizations that you think would be a better home for a dweb/p2p incubation group than a W3C CG?

(Originally published at: https://tantek.com/2019/071/t1/)

[lgrahl]

I was more concerned with the above quoted statement on potential movement to a W3C WG. WICG's charter includes the same statement. I guess incubation could happen anywhere and perhaps WHATWGs suggestion to do this in "a personal document or GitHub repository" might be easiest. 🙂

Note that W3C Community Groups (which this is a proposal for) are open to all (anyone may participate) and without fee (no membership is required) per https://www.w3.org/community/about/

I know but if the earlier suggestion of the charter is being followed, it would eventually migrate to a WG, not a CG.

My only concern was that this could end up in a W3C WG which I would like to avoid for reasons explained above. And by starting a W3C CG it felt as if it would tend to eventually migrate to a W3C WG. If that's not the case - fine by me.