The W3C Credentials Community Group

Meeting Transcriptions and Audio Recordings (2014-today)

Go Back

W3C CCG Weekly Teleconference

Transcript for 2022-10-18

Our Robot Overlords are scribing.
Harrison_Tang: So hello everyone so welcome to the weekly w3c credentials community group meeting so we'll wait for other people to join but before then we can go over the agenda as well as the code of ethics and IP note.
Harrison_Tang: So first of all today's main agenda we invited and Matthew from 3Box Labs to talk about ceramic Network actually a few months ago I remember the community was talking about ceramic and actually one of the community members suggested the idea to invite someone from ceramic to talk about it so that today we're very fortunate to have Matthew taking the time to actually present about ceramic and
Harrison_Tang: what it does.
Harrison_Tang: Any other agendas that people want to bring up?
Harrison_Tang: Alright I'll quickly go over the IP notes first of all just a quick reminder about code of ethics and professional conduct so more or less just be respectful and then make sure that we create a psychologically safe environment where people feel safe to actually voice their opinions anyone can participate in these calls however all substanstive contributions to any CCG work items must be members of the CCG with full
Harrison_Tang: IPR agreement signed so they are links to the emails I’ve
Harrison_Tang: Sent out to the mailing list so just click on that and to join and create a w3c account if you haven't done so these minutes and audio recordings of everything said on this call are archived at the link in the email actually recently we fixed the issue so the issue actually it’ll show up now we're still trying to fix the transcriptions and things like that so that we give us a little time to get all those in
Harrison_Tang: place we use IRC to queue speakers
Harrison_Tang: During the call as well as to take minutes type in Q+ to add yourself to the queue or Q- to remove it and just be brief so that if we have a lot of questions everyone has the chance to chime in now any introductions or reintroductions if you're new to the community or recently new relatively new and you want to reintroduce yourself please unmute
Harrison_Tang: or put yourself in the queue.
Harrison_Tang: All right we’ll also do so at the end of the meeting as well I know some people join usually they would join like five minutes later and things like that so we’ll do the introductions and reintroductions at the end of the meeting as well any announcements or reminders?
Harrison_Tang: Okay and I’ll also queue this up announcement to remind us at the end of this meeting as well.
Harrison_Tang: Any comments on the action items on GitHub?
Harrison_Tang: All right well today we have a relatively light agenda I guess so let's get to the main agenda as mentioned I think a few months ago someone from the community suggested to us to invite someone from ceramic to talk about what does how it intersects with the self Sovereign identity work that we're doing in the community and today we're very very
Harrison_Tang: pleased and want to thank Matthew
Harrison_Tang: To actually give us a presentation on ceramic and I think you'll go over what ceramic is right and then as well as why is it important and then we’ll open up to the general Q&A at the end of his presentation so without further Ado Matthew please take the floor.
Matthew_Davis: Awesome thank you would you guys prefer that I shared slides or do you want me to just talk because I can do either way.
Matthew_Davis: Perfect no problem.
Harrison_Tang: I think if you could share the slides that would be great thanks.
Matthew_Davis: Okay so before I dive super deep into the slides I just want to give you guys a little about me I joined 3Box Labs to work on the ceramic Network about 9 or 10 months ago I come from traditional web2 companies prior to ending up at Ceramic I did a long stint at GitHub building Professional Services for that team so now I'm a developer advocate in this space I've done a lot with data over the
Matthew_Davis: Years I traveled around the world and did a bunch of like database consulting for a while
Matthew_Davis: And that's kind of what led me into the ceramic world of things since ceramic definitely aims to be composable data on like the web3 mindset of building applications so that's a little about me let's dive into what ceramic is.
Matthew_Davis: Also I'm the lead developer Advocate on the team so when we open up the Q&A session there might be a couple questions that I don't have the answers to for sure I mainly interface with developers and integrating this into their applications rather than like high-level strategic view stuff but any type of question that I can't answer for you here I can get over to our leadership team and they'll be able to answer that and I can follow up with you guys so
Matthew_Davis: don't hesitate to try to throw me to
Matthew_Davis: the Wolves but at the same time you guys might get me on one or two so you'll have to forgive me.
Matthew_Davis: So our team's been around for a while. Our core Foundry started at consensus way back in 2016 really focused on building leading web3 developer products most recently we've shipped something around DIDs so decentralized identifiers called DID session we'll talk about a little bit later it's really awesome I think it's like the number one product for our team to be proud about with how they're managing DIDs and sessions for
Matthew_Davis: apps as well as using those
Matthew_Davis: DIDs to control data and provide you know Sovereign data to the owner of the DID and stuff and we shipped compose DB which expands our database layer to allow a lot more indexing to make it more queryable and to conform a little better with traditional standards like graphql so we've shipped a lot of things before that we've worked hard on like chain agnostic capability objects we've worked hard on react libraries like self-id.
Matthew_Davis: and a whole suite of
Matthew_Davis: Developer tools along the way some of you might have actually stumbled upon ceramic and not realized it if you've ever used like 3 ID connect to get decentralized identifier from let's say Meta mask then you used ceramic in some way shape or form that used to be the flagship like product was this 3id connect which would pop an iframe the problem with that is it was very like user heavy the user had to
Matthew_Davis: assign a lot of transactions to do very simple things so it didn't like
Matthew_Davis: Provide the best user experience DID session kind of solves that so I think in order to understand the value prop of ceramic it's really important to walk through a very simple user Journey it's one that we can probably all resonate with everyone of us has signed up to a social app I'm sure whether it be Facebook Twitter LinkedIn right name one you signed up to it excuse me during sign up time you probably answered some standardized information.
Matthew_Davis: Like what your name is, what your birthday might be, where you’re located, a username
Matthew_Davis: And then once you're like logged in maybe you set up site preferences and then you had to go through this really long painful process of finding the people that you like in life to follow or the people that motivate you and you have to like follow them or and then if something goes wrong you have to unfollow them and this isn't a bad flow this is just like using an app but it gets really cumbersome when we go from one app to ten apps or however
Matthew_Davis: many social apps exist a lot of them ask for the same information upon signup
Matthew_Davis: Right whether that's name birthday location we kind of get the point they all kind of had the same flow too once you're signed up and you're logged in then you're still adding and following your friends and your connections and and all the messaging and stuff exists this is not only a problem because we're repeating a bunch of data the bigger problem is that all of this data gets stored on that social app’s servers so if you consider Facebook which I think we can all agree Facebook has the
Matthew_Davis: most wonderful data
Matthew_Davis: Practices of all time uh-huh all that sign up information and all that info about you get stored on Facebook servers and then when we sign up for Twitter it all gets stored on Twitter servers and what we signed up on LinkedIn it all gets sorted Linked In servers and these servers don't communicate with each other and you no longer own that data about yourself you you lose agency so you yourself the owner of the data has no agency over what happens to that data going forward.
Matthew_Davis: And this is kind of scary for developers
Matthew_Davis: too because what if Facebook Twitter LinkedIn whatt if they decide to shut down the API to that data right now a lot of these social platforms they allow open access to APIs so that developers can query the data and use it in other applications but there's still like this really weird like layer of trust that we have to trust that we're getting the correct view back from Facebook and Twitter and whoever might have the agency over this data.
Matthew_Davis: So this is like
Matthew_Davis: The major hurdle that we try to solve and it's not just one that we woke up to one day and said hey like let's let's fix a problem that's not real we're seeing a lot of web3 companies really demand that sovereignty over data exists for their user base and that's really what we're aiming to tackle so for us ceramic what it really is is a decentralized network for composable data and it doesn’t necessarily
Matthew_Davis: Make your application web 3 however it definitely contributes to the web3 stack.
Matthew_Davis: It kind of comes in three very distinct components and if we can wrap our heads around these three components I think what ceramic is becomes really clear at the bottom layer the bare minimum here there's a ceramic node and this is just a small application that interfaces with ipfs it's the underlying infrastructure this is not currently an incentivized layer it will however one day be incentivized so that if you choose to.
Matthew_Davis: just be a node operator you can be a node operator.
Matthew_Davis: So a node is is basically the small application that runs interfaces with ipfs but it is what decentralizes all the data stored on the ceramic Network the purpose of this is when we need to scale we can just horizontally scale by adding another node the main responsibility of a node is really providing the validity of the data so when we request data from a server that that that data is what it's supposed to be that the state is
Matthew_Davis: current there's blockchain proof that accompanies that data to some degree
Matthew_Davis: And says “hey this data is correct” it also really keeps track of the stream history when I say stream I don't mean streaming data like we might find in Kafka or some other big data solution data streams on ceramic are very much just like Json objects of data so you could think like a mongodb document for us we call that a data Stream So these things go through changes consider if a user.
Matthew_Davis: they signed up
Matthew_Davis: to social app a and they said their name was Matt Davis and then all of a sudden I decided to start going by Matthew and I wanted to change my name I'm going to need to update that stream about who I am to say Matthew instead of matt so ceramic will keep track of all of those changes to the data and what happens is when we do a read if it's an unknown read we will rebuild the history of all the changes that have ever happened to that data stream from genesis including
Matthew_Davis: all the blockchain proofs to verify that what we are currently seeing since it was previously unknown to us is valid and accurate to the most current state so the node handles all of this stuff for us which is great because it would be a lot to have to handle on our own so the second layer to that is called compose DB and this is one of our newest products out the door if you used ceramic in the past this might have been called
Matthew_Davis: ceramic we have renamed it to compose DB and added
Matthew_Davis: A bunch of
Matthew_Davis: functionality to it however at this layer this is an API that's always going to be open and that's also what the ceramic nodes help with is maintaining the openness of this API so that we at Ceramic could never take it away from developers like it's always going to be open and available so compose DB allows developers to create and distribute like commonly accepted data schemas for these apps so if we go back and we look at this problem
Matthew_Davis: Of name birthday location this this might be something like a user profile and this is a common schema that's used across apps or site preferences might be a specific schema shape and same for friends and followers so compose DB allows those developers to say hey I'm going to outline what a common basic profile looks like once I build that schema I'm going to publish it to the network and that
Matthew_Davis: allows every other developer in the world to quiery
Matthew_Davis: for that schema and say hey instead of rebuilding a user profile let me just use what already exist and what becomes really cool about this is this is the composability piece if a user.
Matthew_Davis: Creates a basic profile using their decentralized identifier their DID then they can log into any app in the world as long as that app understands how to read the basic profile from ceramic that profile follows the user so they can fill it out in one location and reuse it across all apps so this becomes really useful like if you have a million followers and you want to take that list with you from app to app to
Matthew_Davis: app you could just log in and since that list is
Matthew_Davis: now attached to your DID rather than the application itself that list follows you everywhere and it's just up to the application to read from the ceramic Network rather than you know it managing that entire friends list on its own so compose DB enables that and the top layer here is this data Composites idea and that is where we take the decentralized identifiers and we allow them to control
Matthew_Davis: these schemas that have been composed and built together
Matthew_Davis: So if we go back again and we take a look at this where user signup might be a schema site preferences might be a schema adding a following friends might be a schema we kind of bundle them together into something called a composite and then the app is able to read just the data schemas that it cares about and it does so at this top most level and this is where our DID session Library really comes into play.
Matthew_Davis: this flow looks a lot like a
Matthew_Davis: User goes to an application there's a connect wallet button in that application they click it Metamask pops up if that's their wallet of choice and it says Hey like do you authorize this app to do stuff on your behalf typically those permissions are interacting with this specific data composite user signs that message one time and they get a version of a DID in which now they can the app can
Matthew_Davis: do stuff in a session
Matthew_Davis: Based manner on their behalf for a predefined period of time so the user no longer has to sign multiple messages if they add a friend they don't have to sign a message to make that happen if they change the site Preference they don't have to sign a message to make that happen they basically sign a message when they log in and then the app is allowed to do stuff on their behalf so it's very much a web to kind of cookies experience only built around
Matthew_Davis: DIDs rather than cookies and wallets and stuff so
Matthew_Davis: The crazy thing is every piece of data on ceramic is globally readable which means that we had to come up with a mechanism to protect like malicious rights and stuff and that's where also like decentralized identifiers come into play the way a data stream works is any application that happens to know the schema can read that data but only the individual that did that controls the stream can write.
Matthew_Davis: updates to that data
Matthew_Davis: And this immediately gives us back that sovereignty back to the users so and the agency right so really we're after like this entire New Era of Open Source information we need desiloed data we want data that works across any app in any context regardless of what the tech stack might be on top of that if the community comes together and they just they gain consensus on specific data models like a
Matthew_Davis: basic profile or like
Matthew_Davis: What site preferences look like it encourages a much broader data ecosystem and it encourages those developers to take part and figuring out what that looks like instead of having so many one-off schemas that exist all of this should be accessible from around the web and we really by running nodes and having developers and putting out all three layers of this or all three pillars of what ceramic is we're really aiming to share
Matthew_Davis: the network effects of what’s going on
Matthew_Davis: Between those that are building and those that are innovating and pushing the limits of what web3 data might look like and honestly the thing that got me to join this organization is the user agency Whoever has the DID is the controller of that data and they get to use it how they see fit if I have a basic profile and it's attached to my DID which is may be attached to my Metamask wallet I can update my name whenever I
Matthew_Davis: I want and nobody
Matthew_Davis: can tell me that I can't and that's what's awesome about this and then that's like reflected across all these applications and this like the profile example is definitely arbitrary but I'm thinking bigger like friends lists or there's a lot of thing like my site preferences if I never had to change the color of a website because it automatically switched to dark mode for me I think that just brings a much better user experience to the world for us on top of that
Matthew_Davis: like nobody's gonna like sell my data and that's super important to me and
Matthew_Davis: If they do
Matthew_Davis: Start to sell it then I can just change it or make it unusable for them because I'm the one that has agency over that.
Matthew_Davis: So this ceramic network has a couple pillars these qualities that exist and we had to keep these in mind as we built out this solution first and foremost it needed to be scalable and we achieved that by having a local first design so if I sit down to develop something using ceramic I'm also running a node on my local machine this allows me to have real-time updates during the development process to the data that I'm working with so my app gets
Matthew_Davis: instant feedback.
Matthew_Davis: As I build things on top of that it also increases the number of nodes on the network which helped with the real-time updates for apps that actually run in production the other advantage to this is if you've done web3 stuff which I'm sure most of you have we we know it can be slow because of consensus right when we have to agree that like a proof exists sometimes that takes time but by running a local first design we can skip consensus during the
Matthew_Davis: development time which means
Matthew_Davis: That we’re not constrained to waiting on that to happen for us to see if we're building the right thing in the bigger picture like there is consensus on the network but for development we get to abstract our ways out of that. The second things is that
Matthew_Davis: It had to be mutable like data doesn't make any sense if you don't allow users to change it people change their name frequently that's not an uncommon thing people change their usernames frequently that's not an uncommon thing so it needed to be updatable but we also wanted to maintain every single state that that data stream has ever been in and that was very important that we had that read-only history since that is the only way to prove that the current state of the data is exactly what it's
Matthew_Davis: supposed to be and we do that with built in Anchor commits down to currently ethereum
Matthew_Davis: In the long run it is chain agnostic we do aim to be chain agnostic our development team has an ethereum background so we support ethereum first and then other evm Alternatives and soon hopefully other complete layer ones as time goes on we're also flexible and that means that those data schemas we’re aiming to use graphql and we really give users that single unified interface.
Matthew_Davis: As a developer like all you have to do is understand
Matthew_Davis: how to use the ceramic library and that is going to interoperate with ipfs it's going to get your data anchored down on chain you don't have to learn like 10 different APIs for 10 different apps to make things happen it's one API to rule them all in Lord of the Rings style hopefully when it comes to dealing with data on web3 and interoperability I’ve kind of hinted on this a little bit it's chain agnostic.
Matthew_Davis: We're aiming to support all keys chains and context that we can come
Matthew_Davis: Up with and we do this because this whole decentralized identity piece is built in it is a core fundamental piece to how ceramic Works no stream of data on the ceramic Network can exist without decentralized identifiers they're the only thing that is allowed to control a data stream on ceramic which also means that the only thing that's allowed to create a data stream on ceramic.
Matthew_Davis: So all of that kind of brings itself back full circle to say that ceramic is essentially Dynamic content that's created using flexible data models stored on a decentralized peer-to-peer Network controlled by decentralized IDs and if you can wrap your head around that piece then I think there's a really solid understanding of what ceramic is and what we're aiming to do with it so I think that's the general
Matthew_Davis: Overview I don't want to take up too much time going into specific web3 solving cases here but I would love to open it up to allow for this to become more of a discussion and questions and answer whatever you guys might have I think there's probably value there at that layer at this point right so that's the overview I would love to open it up and talk to you guys.
Harrison_Tang: So Keith you have the floor.
Keith Kowal: Thanks so much Matthew it's super interesting can I ask about I was confused about the Privacy model so I think you talked about how DIDs control rights under the ceramic network but then anyone can read if they understand this schema so if I was to put me let's just talk PII so if I have PII in the ceramic network does that mean that anyone can read it or they can only read it with my permission or can you just talk about the overall privacy model if if it exists.
Matthew_Davis: Yeah this is actually super easy question answer I thank you for asking it so I also think it's a really important one the short answer is there's not a huge privacy model the aim would be to make data open and composable which means that the limitation is that you would not want to store PII on ceramic there's a couple ways that you could but like I don't think anybody at
Matthew_Davis: ceramic would ever recommend doing any of them
Matthew_Davis: One of those ways would be to encrypt your data client-side stored in ceramic and then decrypt it when you need it the problem with that is is now it's not not open so if you encrypt it in application A application B has no way to decrypt that so there are certain things that ceramic is not suitable for things like credit card numbers Social Security numbers any form of PII we would 1 million percent advise to not put on.
Matthew_Davis: All data on Ceramic is not encrypted by default simply because it has that globally read-only mindset to it we want it to be reused we want application developers to not have to design custom schemas and as a result we encourage people to store as much plaintext stuff as we can so it's really useful for user-generated data that isn't sensitive or also like service generated data this is an area we don't talk much about yet because it's kind of like
Matthew_Davis: in Exploration mode microservices often have
Matthew_Davis: needs and they could potentially have DIDs and they could be writing data that isn't sensitive too so things that aren't sensitive is what work really well on data there is a protocol out there called lip protocol and that would be if you did need to encrypt anything on ceramic still don't recommend PII but if for any reason you need to encrypt anything we recommend that you go through lip protocol
Matthew_Davis: to do that encryption.
Keith Kowal: Can I ask a quick follow-up because I'm now a little confused because a lot of the use cases you talked about were things like pro- maybe we're using different definitions of PII because in my definition of PII that includes things like first name last name gender like a lot of the things that would be in a profile when you like sign up to a website so maybe am I missing something here like maybe you don't consider that to be PIIi or?
Matthew_Davis: Well I mean I can that's your birth your name is pretty public right I can look that up on a million forums I like my address is pretty public right like I bought a house that gets that gets published publicly to the city so I don't necessarily consider like my name PII whereas like my social security number my driver's license number even which is still kind of great because I think that's pretty public normally
Matthew_Davis: Like things that could better compromise me might be and I get that you could build like a threat profile like by getting my name birthday and like my pet's name right you could probably social engineer somebody on that front the second piece to your follow-up is the profile thing is very much just a way to for us to talk about composability in a realm that we're all familiar with we had some things using profiles and the advantage to that
Matthew_Davis: is it's all opt-in right so like nobody’s going to force
Matthew_Davis: A profile on a ceramic User it's one of those things if you want to have a profile you can have a profile there's also nothing that stops you from using fake data for whatever it's worth like if you wanted to be somebody completely brand-new in web3 like you kind of could because it's all about your decentralized identifier and not so much the text that you've attached to it so I think there's definitely there might be a small difference in what we're considering PII for sure and then on
Matthew_Davis: Top of that the profile thing is just
Matthew_Davis: something that's easy I think for us to reconcile and a presentation for what could be composable.
Matthew_Davis: Does that answer the question?
Matthew_Davis: I'm happy to go deeper too if need be.
Keith Kowal: Yeah I guess my high-level comment I think only a lot of us look at the who have struggled with PII on blockchains I think that's where a lot of us come from it's not so much about what you think could be PII it's more about the Privacy regulations like GDPR what they consider to be PII and then their rules and regulations around that so I mean that's always kind of my framework for analysis but thank you for your answer.
Matthew_Davis: Yeah that's actually a super good point I'm taking a note of it right now to be honest with you I like to go have like debrief sessions with our engineers and our team after I do these engagements so I took a note of that and because I think you do have a good point right like what we consider to be PII or not might not follow a regulatory standard and that’s something that we should look at so I definitely appreciate you bringing that up.
Harrison_Tang: Oh we have - Nick do you have question?
Matthew_Davis: Yeah that's also a great question thanks for asking that a couple of our Flagship apps that we have on our main net right now are gitcoin so gitcoin just rolled out gitcoin passport all of that is built on top of ceramic metagame is built on top of ceramic dao haus is built on top of ceramic DNS XYZ is built on top of ceramic I think rabbit hole is and there's a handful of other orbus
Matthew_Davis: orbus is a major player.
Matthew_Davis: They’re a major player on ceramic they even rolled out an SDK that further integrates ceramic for social use cases so those are like the main like pioneer web3 apps that exist as far as like some of the data we're seeing hundreds of thousands of DIDs being generated I don't know how accurate that number is it's not hard to go in and create a DID you could I have probably created a good 50 of them myself just
Matthew_Davis: doing testing and application development
Matthew_Davis: So I don't know how accurate that number is or like what it truly represents we are far enough along to where like we're a real product but we're not far enough along to where our metrics have super meaningful analysis behind them we're just now getting to that meaningful analysis behind the metrics that we we do capture so we have a lot of DIDs which would imply that we have a lot of streams which there are quite a bit of streams there's actually a site it's called cerscan it’s c-e-r-s-c-a-n.
Matthew_Davis: I think they're in Explorer for ceramic streams so that's another one they’re kind of a web2 app though I don't think they're deployed in a web3 way but they allow you to explore some of like the data that exists on ceramic and that's kind of exciting if you're into Data as a whole those are our big major players we also have some interesting use cases coming down the pipeline so a lot of those are social a lot of the ones I mentioned are more
Matthew_Davis: social or verified credentials those those kind of exists there we have a couple of projects in
Matthew_Davis: Our pipeline right now though that are branching out to do things like carbon footprints for buildings and putting that on chain or like brand reviews we're working with a project right now their goal is to say solve the fake review thing that like Amazon deals with so if I want to leave a review on my brand new shoes then I have to have some sort of identifier that I own those shoes most likely an NFT and if
Matthew_Davis: I have that then I can write a review and they’re storing
Matthew_Davis: All of their reviews and the validity of those reviews on ceramic or at least that's their their goal long-term they're still really early on so we're starting to see more use cases that aren't social related or aren't verified credential related but currently a lot of it is social and verified credential which kind of throws back to the whole black basic profile idea that's something thate our current Builders can reconcile with a lot so
Matthew_Davis: we'll probably get better examples of composability as we get
Matthew_Davis: Different use cases built out on ceramic there's a good handful of applications on our main net and that does not account for all the applications that come out of hackathons at like ethglobal and stuff that are running on test net and I don't have the test net numbers.
Harrison_Tang: Hey Matthew I have a question in regards to like Speed and Performance because when I last check ipfs which ceramic is built on top of it's a little bit slow so when the use case that you're talking about signing in usually people have service level expectations in regards to like almost immediate right so how what's the kind of the performance here and then how do you kind of if you have solved the performance issues
Harrison_Tang: how do you solve it?
Matthew_Davis: Wonderful question I love it that there's a lot to unpack here so I'm going to talk about two different things because they're both performance-related the first is the sign in piece because I agree there's like a service level expectation that if I click the sign in button I need to be signed in right now I don't have time to wait web2 has done a good job at getting us signed in quickly for the most part currently with our new session library that we shipped
Matthew_Davis: that wraps around DIDs signing in is really painless it's really fast because it is not coupled with ipfs so all the DID stuff doesn't actually write down into ipfs so I could go create a bunch of decentralized identifiers right now and they're not anchored into ipfs so we abstract the fact that ipfs is a little bit slow away from that what ends up happening is like when that DID decides to
Matthew_Davis: generate data there’s a
<keith_kowal> If DID docs are not on IPFS - where are they written? A more traditional database?
Matthew_Davis: Field in the data object that says controller and we put the DID in there for reference and that's who can control the data so now we do some logic that says hey there's an update that wants to go into this stream is it from the DID yes it's from the DID okay we'll make that update so all the sign-in piece is really fast it's a actually we implemented and helped Pioneer the sign-in with ethereum standard so if you're used to sign in with ethereum or have any experience with it.
Matthew_Davis: it's as fast as that
Matthew_Davis: You click the connect wallet button you select metamask metamask pops up says sign this message you click sign and you are you have a DID and you're ready to go in a matter of however fast you can click those buttons so it the sign in piece is pretty immediate it's really actually a wonderful user experience for web3 I remember when I was first getting in the web3 the whole how do I connect to this app was was kind of a painful process this is actually
Matthew_Davis: I hate using the word elegant when it comes to code because it’s really
Matthew_Davis: Hard to have Elegant solutions for code but this one really feels elegant it's also extensible as I said right now it's really built on top of this sign with ethereum standard we are currently in talks with Solana and pioneering what the sign in with Solana standard is going to look like so hopefully we can Implement that chain agnostic you know DIDs here pretty soon and sign-in Methods at least for your apps so from that point of view we're talking pretty immediate.
Matthew_Davis: going down into the ipfs layer which is where the data is right so ceramic.
Matthew_Davis: Sits right on top of ipfs and it's an interface down to ipfs we solve this problem because of the local nature of a node I don't have a diagram but I'll try to talk this through so you have an ipfs node we’lll call that the bottom layer so layer 1 or layer 0 depending on how you want to count will be in ipfs node and you're right like that's connected to an ipfs
Matthew_Davis: swarm and it's slow to propagate
Matthew_Davis: And that is the nature of ipfs on top of that is ceramic and when we make a call to data ceramic holds the most recent change so the best way I can explain this is to think of the command line tool git or just any Version Control for code when we do a git init and we get that first commit in there so we'll git commit initial commit or whatever right we have that first Genesis commit.
Matthew_Davis: And every commit we make we’re kind of like tracking like what's different like what's changed between commits ceramic streams on the same those Genesis commit and that lives on the ceramic node initially and then if a user makes an update they're basically making git commits to some degree to a ceramic stream now they're not git commits but they're similar we're committing the changes to a ceramic stream what's happening is the ceramic node is always in the know of the most
Matthew_Davis: current state so the tip it knows the latest commit the previous commits are what gets sent down to ipfs so all the historical commits end up on ipfs so if an application that is tightly coupled with a ceramic node needs data it is asking the ceramic node for the tip and it's getting it back instantly it's as long as standard HTTP request takes so we're talking milliseconds it's getting that information.
Matthew_Davis: back instantly if an application
Matthew_Davis: that is not tightly tightly coupled with a node requests data then the node that has the tip will send the tip over or a copy of the tip over to the new node and the node for application 2 will query ipfs and will rebuild the data from Genesis commit all the way to tip and check all the blockchain proof along the way so the very first time a brand new application
Matthew_Davis: requests information
Matthew_Davis: That it doesn't know about it can be a little slow on query but from there on out that node will also keep track of the tip so there's like lib PDP pub/sub messaging that takes place between the nodes to keep the tips in sync so it will know about the tip and all subsequent reads will be much faster so the performance depends on the situation how we solve the the slow ipfs
Matthew_Davis: part is by
Matthew_Davis: Having those nodes keep track of the most current state and we only lean on ipfs when we need proof to verify that that the the current state is accurate there's a big consensus mechanism that runs too in case like we need to like there's two rights that happen at once it was two reads that happen at once the nodes will communicate with one another to look at like how everything got anchored down to chain through ipfs and then they'll decide who wins.
Matthew_Davis: Thats a very well documented thing though in the docks and I'm going to mess it up if I tried to tell it to you completely but that's just know that there is a consensus mechanism to solve for that but that's how we get around like those performance implications of ipfs.
Harrison_Tang: Thank you thank you Matthew and Keith I think you have a question earlier in the chat right.
Keith Kowal: I think I answered my own question so I add asked where the DID docs were stored but I think if you're using DID pkh then I guess they're just instantly composed sorry correct me if I’m wrong my reading my documentation is that you're using did TKH so you're not storing did docs you're just composing them on the fly when you have those kind of sign in applications would that be correct?
Matthew_Davis: Yeah they get composed on the fly so when I when I connect my wallet I get a standard DID I forget the method that's used on the wallet right away then that gets turned into DID pkh the DID pkh is passed over to the app obviously that's the public key hash for it that's what manages the session but the DID itself like we know of DIDs because it's like we can tell how many digits are created not sure
Matthew_Davis: exactly how that communication with
Matthew_Davis: The data takes place that's a good question I would love to follow up with a much better answer for you however I know what I do know is that I can create many DIDs via command line tools and other stuff that aren't pkh DIDs and still use our libraries so those aren't stored either like if I were to sit down right now at my terminal and created a DID using DID key that's not like stored on ceramic anywhere.
Matthew_Davis: It is if I publish it to the node but like I don't know if the DID session which is what is responsible for DID pkh I don't know if that inherently publishes it to the node that'll be something I can follow up on with you though to get you a better answer.
Harrison_Tang: And Matthew a different question so earlier if I remember correctly you mentioned that there are some folks actually using ceramic to kind of actually store verifiable credentials using ceramic so if that's correct like can you kind of share with us what are some of the use cases because you also mentioned that it is advised not to put really
Harrison_Tang: sensitive personal information.
Harrison_Tang: on ceramic Network so what kind of verifiable credential use cases are people using right now on top of ceramic?
Matthew_Davis: Yeah great question.
Matthew_Davis: There's a couple different use cases I think I think gitcoin turns these into VC's so gitcoin passport is the number one thing I could point you to for this so what they do is they allow you to build like a more robust like profile around who you are so if I if I create my gitcoin passport I can get stamps which I'm pretty sure they're storing as like VCS in those stamps are like hey I'm going to I'm going to verify my Twitter account into this passport or
Matthew_Davis: I'm going to verify my Facebook account so they’re building kind
Matthew_Davis: Of a link between web2 applications so that they can be reflected better on web3 apps they’re right now like the biggest player in any type of like VC kind of space for for what ending up on ceramic I am 99% positive all those stamps are VCS I did not work super close with gitcoin when they did the integration though so you should probably do due diligence to find out if they are VCs but I'm pretty positive those are.
Matthew_Davis: then we actually jus had a
Matthew_Davis: Chat with an organization that's doing something super unique over at devcon last week they are building a know your developer platform so one of the things they're doing is looking at historical like GitHub behavior from a developer and building like know your developer kind of credential that they will eventually want to store on ceramic but they're not as polished as gi coin is yet they don't have that out even on test net that's something that they're aiming to do so we're
Matthew_Davis: seeing things that and so that one the whole GitHub one
Matthew_Davis: Is is all public data right like GitHub has the APIs for querying everything you've ever done as a developer that you've done in public and so they're using public data that's already available so it's not anything the user didn't opt in to being public as is and then they are actually scrubbing a lot of the the PII out of it so they don't care about your name they don't care about your email they’re scrubbing that out because the GitHub API returns that anyway.
Matthew_Davis: every time you run a query on like a specific user
Matthew_Davis: Or a repo same with gitcoin I don't know that they're scrubbing it necessarily however it's a 100% opt in situation like I don't have to connect my Facebook to my gitcoin passport if I don't want to and I think that's largely the mindset of how it's being handled right now it's a 100% opt-in situation at the moment.
Harrison_Tang: Thank you any other questions for Matthew from the community.
Harrison_Tang: All right Matthew I have one last question for you so so like how how do you think like ceramic can help increase the market adoptions of self sovereign identity because you know as we know self sovereign identity even though we are in the community and it seems like it's getting more and more popular but it's still a long ways to go in terms of getting mainstream adoption so how how do you see like ceramic or from your experience is the
Harrison_Tang: the intersection between web3 and then the
Harrison_Tang: self sovereign identity community how do we actually work together to increase the market options of SSI.
Matthew_Davis: That's another wonderful question from Ceramic’s point of view we are pioneering the use of DID PKH and we really think that that ties into self Sovereign identities very tightly from here on out every DID that is created on ceramic that that we encourage the creation of is pointed to using this DID pkh method and what that actually does is
Matthew_Davis: it takes all
Matthew_Davis: block chain wallets and it kind of turns them into DID wallets right any wallet can now become a DID and that gives you all the sovereignty over your identity there so we are by I think by driving adoption of of using DID pkh through various hackathons and just publishing a really well thought out and well built package that's easier for developers to use I think we definitely push things forward when
Matthew_Davis: it comes to using that as.
Matthew_Davis: an identity mechanism on top of that our like our CTO not to Humble brag too much but like he's very active in the the DID standard itself for for W3C so if you go look up the DID standard Joel Thorstenson is our CTO like he's he's an active author for what decentralised identifiers look like so from just like the the point of view of.
Matthew_Davis: values to some degree as like
Matthew_Davis: Company values I don't think there's a bigger commitment to anything in Web3 outside of what we have for identity the identity piece is critical for us having Sovereign data the identity piece is critical for user adoption especially as we've seen web2 kind of violate privacy along the way so from for us like I think it's very a very obvious flag planted in the ground that identity.
Matthew_Davis: is key and it's important and we're going to continue
Matthew_Davis: pioneering what that means to make it easy to onboard users into as I said about like DID pkh it couldn't be easier to sign into a web3 application right now like sign in with etherium was great and then we're automatically you know rolling DIDs when you use that standard and you have that and it follows you everywhere because it's attached to your wallet we're also working right now on figuring out how to do like multi DID so if I had multiple wallets but I still kind of want them to
Matthew_Davis: coupled with a bigger ID like how does that work and that's something that’s in like the
Matthew_Davis: The scoping phase for us so I wouldn't expect it any time soon but it is on our plate so we're going to continue pioneering ways to reduce the friction of the web3 user getting a DID while making it very obvious that the web3 user is using DIDs rather than a direct wallet address and we're going to continue doing that in the ways that we've gotten this far which is hackathons and good packages and just encouraging the community to build with
Matthew_Davis: these standards we're also going to continue contributing
Matthew_Davis: back to those standards that do exist around decentralized identity that's not a one and done for us we're fully committed to having Sovereign identities on web3.
Harrison_Tang: Thank you thank you Matthew Si you have the floor.
Matthew_Davis: Yeah that's a super solid question I actually just got out of a huge dkp like presentation at dev con last week there is definitely a use case for this ceramic itself is probably not going to Pioneer what that looks like though there's nothing that would prevent you from integrating zero knowledge proof into the way your application works and then storing whatever you need to on ceramic but I think we're at this stage
Matthew_Davis: we’re largely leaving that in the hands of the community to solve.
Matthew_Davis: We're really focused on just making sure we put out a good product for example like indexing data with something we didn't have a month ago and now we do have it so our focus is largely on queer ability and interactions with the data but for like that type of stuff we are absolutely encouraging running zero-knowledge protocols along with ceramic to figure that out we're super super stoked to see what the community does on that but we're not
Matthew_Davis: at a point where we can put our own engineering cycles into the
Matthew_Davis: because we just don't have the bandwidth.
Harrison_Tang: All right I just want to thank you again Matthew for taking the time to present ceramic to the community appreciate the time and appreciate ceramic like committing so much into Open Standards.
Matthew_Davis: Thanks for having me I love talking about this stuff and it's always great to be in a group that asked very thoughtful questions and is very clearly thoughtful about what web3 means to the world so I just genuinely thank you for having us here today.
Harrison_Tang: Great great and then last announcements or reminders on the community?
Harrison_Tang: Last introductions and reintroductions?
Harrison_Tang: All right thank you we're right on time thanks thanks a lot and thank you Matthew again thanks.