Should DID Methods be hosted on W3C?

[ydennisy]

Hello Manu & Dave, My questions is should various DID methods be hosted on the W3C website - which makes them a standard? My feeling is that a method is an implementation of the standards. Best, Dennis

[msporny]

My questions is should various DID methods be hosted on the W3C website - which makes them a standard?

Eventually, yes. For now, no. For a spec to be hosted on W3C as a standard, it needs to go through the entire W3C WG process and none of the DID Method specs have gone through this process. We are a very long way from that happening... it requires a fairly significant effort from the industry. Veres One is the only spec at present in W3C format and we'd need multiple other companies to back the spec and take it through the W3C process (which is a minimum $1M+ investment from industry).

So, for now, we need more implementation experience and adoption before we go that route... but do expect it to happen in time. Check back in 3-4 years. :)

[ydennisy]

@msporny thanks for your reply.

So my question stemmed from the fact that the DID method spec for Veres One is being hosted on what looks like the official W3C site. Yet this is a commercial implementation, to me this does not seem right.

Also who allows for others to add their method specs to the W3C site?

[msporny]

So my question stemmed from the fact that the DID method spec for Veres One is being hosted on what looks like the official W3C site. Yet this is a commercial implementation, to me this does not seem right.

To be clear, the spec is hosted on the W3C Credentials Community Group site, not the official W3C site.

We have handed over all intellectual property for the Veres One DID Method spec (patent-free / royalty-free) to W3C per the W3C Community Group process. The source code is licensed for use by the Veres One project (effectively AGPLv3 with a "for profit organizations can't fork clause"). So your "commercial implementation" assertion confuses me.

Also who allows for others to add their method specs to the W3C site?

The process for doing so is to join the W3C Credentials Community Group, convince the group to take it on as a work item, then publish via the W3C CCG. That's what we did. The process is open to the public and anyone can do it.

[ydennisy]

Firstly, I would like to say that this is not a personal attack and I am just being inquisitive and also taking the viewpoint of what I feel the many would see.

• Whilst it may not be the official site yet, the idea is that it moves to it at some point. Also it does look quite official with W3C branding to most people.
• I understand you have released many components open source, however the business will be looking to make money in some way (V1 or any other related entity) through the implementation of what is being worked on as a spec, to me this seems like a conflict of interest.
• Whilst the process is technically open, if let's say (hypothetical), I create a new method and want to add it to the spec, but it is competitive to V1 or SOV etc, it feels like it is quite easy for the leaders of this group to block this with some excuses.

Again I am not saying, "you are bad people, and you will block others" but to me it seems like it is a possibility from a theoretical perspective.

Dennis

[msporny]

We appreciate the feedback, are not taking it personally, and think that this discussion is healthy for the community.

@ydennisy wrote:

Whilst it may not be the official site yet, the idea is that it moves to it at some point. Also it does look quite official with W3C branding to most people.

We are using the template that is provided by W3C, it would be against W3C Process to use any other template. In other words, we're doing exactly what W3C asks of Community Groups. I understand that an outsider may misinterpret the specification as a W3C specification and the W3C is working on different stylesheets to try and correct that misunderstanding. That said, that is out of our hands and is not within our authority to change.

I understand you have released many components open source, however the business will be looking to make money in some way (V1 or any other related entity) through the implementation of what is being worked on as a spec, to me this seems like a conflict of interest.

We are transparent on what the funding structure looks like:

https://veres.one/network/economics/

If you want to see the proposed breakdown, it is here:

https://veres.one/network/funding/

The other miscommunication may be that Digital Bazaar stands to gain a lot of money from this endeavor. We do not -- 5% of gross revenue capped at $25M/year. No other project has a cap that low for the people that created and launched the technology. Other caps are in the hundreds of millions if not billions of dollars. The people that run the ledger are not Digital Bazaar, it's the Veres One Project with an independent Board of Governors and a public and transparent community... 95% of the funding goes to running the public utility.

Fundamentally, the network needs money to operate and we're fully transparent about where we expect that money to come from.

Whilst the process is technically open, if let's say (hypothetical), I create a new method and want to add it to the spec, but it is competitive to V1 or SOV etc, it feels like it is quite easy for the leaders of this group to block this with some excuses.

Let's assume your hypothetical situation... we block the addition of your DID method to the registry. Nothing stops you from claiming the DID Method anyway and gaining traction... the DID Method registry is non-normative.

https://w3c-ccg.github.io/did-method-registry/

You don't even have to register your DID Method if you don't want to. Doing so is a good idea, though, so others know that your DID Method exists and can go to the spec that defines how it works.

The community has never blocked a request to add a new DID Method to the registry. If it does, it'll happen in the public and the public will lose trust in the community. Everything the community does is on public mailing lists, public teleconferences with fully recorded audio and transcriptions, and source code issue trackers. If such a thing happens, it will be a very public and will result in disastrous negative publicity for the group. All of the incentives are to NOT do something like what you're saying and we specifically designed the governance of standards community to prevent such a thing from happening.

[ydennisy]

Thanks Manu for your replies.

I am still deciding if Orbis should build or use a DID method...

Do you have any insight into that decision?

[msporny]

I am still deciding if Orbis should build or use a DID method... Do you have any insight into that decision?

Having been involved in a company that had to build a DID Method from scratch, don't do it unless you absolutely have to. It is a very expensive endeavor both technically and from a governance perspective.

The only time you'd want to do something like that would be if none of the existing DID Ledgers supported the sort of thing you wanted to do. I'd start out by using any other DID Method that supports most of your use cases, see if that works, and only if it doesn't work do you implement a new DID Method.