- From: Manu Sporny <msporny@digitalbazaar.com>
- Date: Thu, 14 Oct 2021 18:08:39 -0400
- To: "W3C Credentials CG (Public List)" <public-credentials@w3.org>
On 10/14/21 5:48 PM, Nikos Fotiou wrote: >> So the Ethereum community will have a choice to consolidate to one >> method that's backed by the Ethereum blockchain, or lose out to other >> blockchain networks that do consolidate. > > But this doesn't show luck of interoperability? An Ethereum resolver > should be able to accommodate all of them, may be with some pieces of > configuration for each of them Well yes, it should, it's just that no one has built that for Ethereum yet... and we expect them to. Ah, I think I know where we might be miscommunicating... and that is in how W3C defines "interoperability" and what the minimum bar is. Typically, the minimum bar for a W3C specification is to demonstrate that one specification has been implemented by at least two different implementers. The DID Core specification is a data model specification, so the general approach there (which has worked for many W3C data model specifications) is to demonstrate that you can take the data model, serialize to a serialization, and then go back to the data model. The DID Core specification has 42 implementations that did that (yes, that is a ridiculous number of implementations). Some of the objector's arguments seem to be "yeah, but you didn't show DID Method interop". It's a fair point, but the DID WG Charter forbids us on working on that because people were afraid that a big company would jump in and try to get their centralized DID Method "ratified by W3C". So, the negotiations resulted in NOT do DID Method interop and put it out of scope. However, we did prove DID Method interop outside of W3C -- at plugfests that are ongoing in US, Canada, and EU. Demonstrating interoperability for HTTP wasn't "demonstrate that ALL websites in the world work", it was "demonstrate that two web servers can be implemented to speak the same protocol". > Check institutional login here https://ieeexplore.ieee.org/Xplore/home.jsp > You can login using the user management system of over 100 universities. > Each university uses its own type of "registry" and you need only one > resolver (may be with a configuration file with over 100 entries, but > still). Yes, and university federated login is an example of scaling failures wrt. federated login. For example, I put in my local (fairly large - 40,000 students and faculty) university and got nothing that popped up. That could be equated to a single DID Method (the way it does resolution is exactly the same). That said, there are systems being created that help websites resolve multiple different DIDs. Markus' Universal Resolver is one such system: https://dev.uniresolver.io/ So, the website would either just use the default DID Methods (which have been vetted) or add new methods via an allow list. I'll note that the Universal Resolver currently supports 36 different DID Methods, so websites don't have to care about the complexities of each DID Method system... they just list the ones they accept. > So, IMHO saying that "we are targeting interoperability and convergence but > we are missing some parts" is an honest response to Mozilla's and others > comment. Well, the argument is over "what is interoperability" -- which we defined in the WG during the Candidate Recommendation phase and no one objected to the way we were going to demonstrate interoperability. Mozilla's argument is that we didn't achieve the kind of interoperability that matters and there is a debate over 1) whether or not we were allowed to do that in the charter (we weren't), and 2) whether or not we accomplished it outside of the charter (we did), and 3) whether or not Mozilla believes that (they don't seem to, so far). > But Evernym's blog post appears to me like saying "we already have it and > you are stupid you didn't read the specifications". *lol*, if that's the message you get, then we'll want to correct it. Drummond did write a very good response on the AC mailing list covering exactly that concern over the blog post. -- manu -- Manu Sporny - https://www.linkedin.com/in/manusporny/ Founder/CEO - Digital Bazaar, Inc. News: Digital Bazaar Announces New Case Studies (2021) https://www.digitalbazaar.com/
Received on Thursday, 14 October 2021 22:09:00 UTC