Hi ! 👋
Running full speed ahead, DIN is charging many angles of information sharing and projects to get SSI into the wild. More info about this in future newsletters. Let's get to our main question.
If everyone shall have a wallet, how are all these going to talk to each other?
Interoperability is the panacea of digital identity. We have OAuth2 and all the technologies the big tech are relying on, and it works to make the lives easier for users, but it does feed the big tech with more data? This interoperates in the sense that it is easy to implement towards someones OAuth endpoint and use their data as a trusted provider. But I cannot use Google data inside Facebook and other big services. And I don't want to give these big tech insight to every online service I use. The technology is not interoperating at the right layer.
DIF puts it well, "Don’t we already have a standard? What’s so hard about interoperating?". Standards are great, but there are two more points to cover, ref. the header picture. And even EU and AU have not talked about what standards they are using which will make this interoperability case even more difficult, ref our last newsletter, and our last event.
A standard does not solve the interoperability problem on its own. It can solve lots of problems but not interoperability alone. TCP is a standard, but it was all the work around to get this working across a whole layer, that made it interoperable and allowed for more services on top! Agreements and specifications. There are vizualisations to showcase what layers that needs to be addressed and what particular technologies are adressing the different layers. And there are multiple projects and working groups looking to get the right discussions going around the globe, such as DIF and Evernym. Software in itself does not collaborate with other random software unless there is a standard, specification and agreement able to link them.
So, what paths exist to make sure every wallet can be talking to each other? We can follow the vizualisation mentioned earlier and point to some important initiatives.
The two most obvius
VC & VP - This is very obvious and is one of the most versatile credential formats we have today, based on the level of work that has gone into it. This will carry the data and identity attributes in a signed, trusted and semantic ready format
DID - We have an identifier problem in the world today, and this solves it. Combined with a universal resolver. It speaks for itself.
The lesser known
DIDCom - In short, this will make sure that the message between the wallets is in a well known envelop to pass along whatever transport at hand. Go into more detail about DIDCom.
JSON-LD semantics - If you want to play around with the standard, there is a playground. Semantics are important because the user doesn't know what data or attributes the other user have in their wallet. All the user knows is that the user can ask for a known Drivers License, and because of JSON-LD and the schemas, the users wallet can know what data to expect and know what to ask for. This abstracts the API documentation layer, and relies on the semantics of the expected data that has been built up of schemas!
PE(Presentation Exchange) - It can be hard to ask the right questions about this complex data structure. For a wallet to be able to ask another wallet something, they first do a DIDCom message to connect to the other wallet, and the DIDCom message can have a presentation exchange data blob in it, that allows the wallet to understand the query of the other wallet. This presentation exchange is semantically correct, meaning that it will rely on the VC and JSON-LD semantics to query for data, and not take any assumptions on what data is available.
If all these specs with reference implementations and tools, where abundant, I believe interoperability would have been a piece of cake. Everyone would agree that this was the panacea, and we could live a happy digital life. But the standards work is heavy, building the tools in all different languages possible is heavy, and also building supportive tools on top is heavy. That is three heavy tasks, not something that a single person or company can do!
Join in on the interoperability talks at https://identity.foundation/interop/ or just start talking about it to make interoperability something really really important!
Other news
