April Evangelist Call - 1100 EDT edition

Alan: Let's get started
… We record these

Opening comments by Alan

Alan: Thank you everyone for joining. This is the 11:00am EDT edition of the monthly Evangelist call
… We had about the same number of people on the 6:00am EDT call today
… Ivan did a great job
… This will probably be our last call on Webex. Will shift to Zoom for the May call.
… I hope everyone is ok. Your family, colleagues in this time of COVID
… We are seeing people interested in joining W3C in several areas. That is encouraging. The work is continuing obviously.
… The Advisory Committee meeting will be virtual.
… There will be video presentations, and then live Q&A event on 18-19 May.
… I am looking into getting you passes for the recorded stuff and the AC meeting.
… Bill K will have access, but I will take care of that this week to get you added in.
… We are still hopeful that TPAC will be held in Vancouver the end of October.
… Karen and I had a conversation with some Vancouver-based members and it may need to be a hybrid.
… BigContacts updates
… will put those on the Evangelist operations page
… Seeing no one on the queue

DIDs Presentation by Ivan

Alan: I would like to turn the floor over to Ivan Herman
… He is the lead on the DIDs work
… Will turn over to you to screen share

Ivan: Good morning to you
… I hear an echo

Ivan: now it's fine
… Everyone can see the presentation?

Alan: yes
… Good morning, everyone
… Decentralized Idenfiers (DIDs) is what I will talk about
… and I will reference it as DIDs because it's a mouthful
… Here is the url to the slides
… First, an introduction of where all this work comes from
… I think it's relatively clear to all that identifiers play an enormous role in the digital world
… Important to identify things on the web or around the web
… You need it for reasoning, associations, control, anything computers do with things, persons and concepts
… So the digital economy relies on a proper identification to combine resources
… Identifiers are not only web sites...just about anything...[reads list on slide]
… and of course WoT should all be identified
… But of course there are identifiers around already; so what's the problem
… A few years ago I was a researcher
… If you look at the two scholarly references, you see my name in these two references
… But if you know me, you know I am not author of the first one, but author of the second one.
… So and identifier must be able to separate homonyms
… In scholarly world a clear identification of person and the work is essential bread and butter
… I have multiple identifiers

[slide shows list]
… ORCID is used for scholars
… but you may also want to separate your private life from professional life
… So are the identifiers the right ones?
… There are a number of problems with current identifiers
… Identifiers should be easy to create
… But a web site depends upon buying a host name
… Email address when used as identifier is not cheap
… it should be decentralized, but not true for emails; domain or hosting could disappear
… Persistent
… Using my w3.org address; this could disappear after I leave
… or my personal domain disappears if I don't pay for it
… To know how ORCID works, you can create a URL, but that is difficult
… Identifiers should be cryptographically verifiable
… a dot com address could be misleading
… this example of ivan-herman.com is not me
… If you look at these identifiers
… https URIs are resolvable, but are centralized, non-persistent and complex to create
… ORCID are persistent, verifiable, and centralized, but resolvable by a special URL
… the UUID from my own name
… is persistent, decentralized, easy to create, but NOT resolvable or verifiable
… So the point is no identifiers display all of the characteristics desired

[repeats]
… Machines can somehow prove a specific subject and this is important
… The goal of DID is to define a self-sovereign identity
… This is what the WG is doing
… Let me define a high level view
… DIDs and DID Documents
… Left side of chart is DID and there is DID Document
… contains cryptographic materials, the essential thing
… after some crypto verification
… it can go back to whoever wants to check
… this is the round that happens
… Global, Distributed, key-value database
… meta is used here to differentiate between different data bases
… used for different approaches
… different methods can have different characterisitics
… may be based on distributed ledgers (i.e. block chains)
… DID docs can be stored on github
… or there may be ephemeral DIDs on intelligent devices
… what method is implemented depends upon the requirements of how it is used
… the DID documents
… main point is to contain cryptographic information, to verify
… the trusted comms or the controller
… It may also include additional information such as references to web sites
… It may exist as a blockchain
… in some other cases it may be generated on the fly
… I hope it's clear that DIDs and DID Documents are closely coupled
… DIDs cannot be used without DID Documents
… DID+DID Documents make use for various purposes
… as a decentralized crypto keychain
… PGP, key server somewhere
… one at MIT

maintained centrally
… and have a decentralized version
… these docs can be serialized
… as JSON, JSON-LD and CBOR
… Some use cases
… to put you in context
… there are some obvious ones
… securing unique identifiers for people, objects, things
… literature references, identify things
… because they are crypto secure, they are probably more secure
… If you know what Verifiable Credentials are, they use DIDs
… discussions to use combination of DID + VC as a tool for the COVID-19 indexes; help to ensure privacy of information
… this can be a good tool
… If there are new pandemics, they might come into play
… Another application that is under deployment at big department stores
… is identification of goods
… Get the DID on your device
… device can go resolve it to a DID Document
… check whether DID for the camera is really the one being controlled by the right organization
… and may provide further information about the camera in this case
… Very important that the DID is unique and persistent; where the crypto comes in
… not possible to hack information
… other thing is that you can create DIDs in large numbers
… you can have each camera with its own DID and get the information
… a number of dept stores are looking at this
… I don't think I have authorization to reveal who they are
… Another application area is the pool of relationships
… a number of people get together to work on a project
… they may not know one another but want to exchange encrypted information with each other
… This is where DID comes in
… Person A sends to Person B
… person B can retried DID Doc belonging to A
… she can encrypt her data using the keys of the original guy
… simply an application of crypto where keys can be found easily
… the DID method can store this information
… Short technical dive into DIDs
… The DIDs themselves
… in the world of Identifiers and world of IETF
… there is concept of URI
… and different sub-parts
… the best known are the URLs
… mostly HTTP
… or you may remember FTP
… all URLs
… but there are others: URNs
… to express ISBN or ISO docs
… and some others like INFO
… some schemes that exist
… DID is a new type of URI
… it follows the URI specification in terms of syntax
… begins with "DID" string
… then you have a name to identify the Method
… the distributed ledger and its resolution
… then there is an identifier
… a long list of characters and numbers that is an identifier that is dependent on the @
… DID creates new URI, with sub areas
… part of same URI world as defined by IETF
… why it's important that it's a URL
… that means that a large infrastructure already out there that uses URIs can use DIDs
… an a element defined in terms of URIs
… put DID; or in an RDF triple
… so I can put a URI and it is a perfectly valid RDF statement; so DIDs are part of the web
… of course you need a resolution somewhere
… Here are some examples of ledger-based DIDs
… DIDs built on top of the Bitcoin BLockchain
… specific identifier is generated from the Bitcoin transaction position
… ids a position on the blockchain
… also ledgers set up specifically for this purpose
… Sovrin Foundation set up a ledger
… for DIDs
… there are other methods based on Ethereum; others like Veres One
… expect others to come up
… they are meant for a general solution
… On long term, one will emerge as more permanent
… There are others not based on ledgers
… Key is one example
… important point is it is ephemeral
… if you want to change the keys, you throw it away and generate a new one
… here you don't care, so it's for things like IoT
… Another one called "Peer"
… it stores keys in github
… and used for persons who want to talk to one another
… and there are other methods being used
… Lots of experimentation is happening
… to explore different methods
… quite active development work
… and we will see a convergence of methods
… A few words about DID Documents
… there is an abstract model
… DID Document is related to the DID subject
… includes separate DID for the controller
… if subject is my dog, I am the controller
… it expresses the public crypto keys
… may include separate proof section
… DID Doc is serialized in JSON, JSON-LD or CBOR
… DID Doc can be converted from one version to another
… CBOR can express things that JSON cannot
… Showing JSON DID Doc structure [slide shows code]
… ID for subject; controller; publickey, authn; proof; service
… Public keys are important; they identify the subject
… here is an example for an RSA key
… another for an elyptic curve based key
… can be used for applications I referred to
… or for closed purposes
… separate area for authentication
… can refer to a key
… keys listed in authn must only be used for authentication of the subject
… then you can have a proof; signature
… have to ID signature type
… and some metadata
… how signature looks depends on the type
… There are currently two documents from the WG that have been published

Alan: If there are any questions
… please ask

John Simmons: Great presentation
… What is relationship to federated identifies?

Ivan: You mean the one MS started a while ago?

John: yes, Tony Nadlin and Mike Jones

Ivan: Mike Jones is a member of our WG

John: FIDO

Ivan: we have some sort of arrangement with them already, so we try to be in the same boat
… They are working to start a project/group
… on encrypted data models
… they would use DIDs
… there are a lot of people doing both
… in our WG
… @ is a core figure
… Daniel Ducher

John: yes, I know this name

Ivan: there are relationships
… other org that has a strong relationship to what we do is Hyperledger project, hosted by IBM
… one of Hyperledger projects will be a core implementation example of DID
… you had two questions

John: yes, when talking about people v things
… often times its the ability to prove you have an attestation
… like a driver's license
… is that related?

Ivan: for example the proof that you have a driver's license is under the remit of Verificable Credentials
… even the people working on DIDs and VC are the same people
… the owner of the credential is identified by a DID
… DID is never used alone
… have a strong relationship
… Do you know Kim Cameron?
… he is an identity person

Karen: I wanted to respond to the Kim Cameron question.

Alan: Any other questions?

Closing remarks by Alan

< Ivan: I know some people on this call

Alan: yes, she [Eva Virag] is very Hungarian