 Hey, good day folks everybody Hello, Argentina one Really happy Super great Hey, Dimitri doing good Dimitri, do I know you? Good Dimitri, would you like to introduce yourself? Okay Okay Okay, let's see Our Documentation for today Feel free to add yourself to the attendees list. Uh, this is the aries did com v2 working group for uh, december 19th 2022 we meet weekly and Um, I should remind you about the hyper ledger antitrust policy, uh, and the hyper ledger code of conduct and Yeah, feel free to add yourself to the attendees list and Who else anybody else? Uh, no, let's see. We've got bruce and rhodo and hakan and judith So, uh, looks like no one else is new Great, okay. Um, let's see quick, um updates I am don't know of anything with aries agent test harness. Uh, I don't know Dimitri, are you interested in did com v2? Let's see here. I'll put you on mute If I can let's see Give me one second to Claim ownership Yeah, give me give me one second. I'll I'll claim host and then we can get started Okay, okay, I have the power now Do we need to record the meetings? Uh, no it, uh, I believe it auto records so though I can double check that Yeah, I see the rec button The recording, yeah Yeah, it is recording I think the record them got post this like not to this somewhere. I don't know Okay, I have ejected dimitri. Sorry for that Uh, let's go back to sharing That's the difficulty. I guess of an open meeting Okay, and of course Now my go okay and my screen has changed All right Yeah, sorry about that everybody Yes, we are recording. I think that that happens automatically for us All right updates. So aries agent test harness. I haven't done anything with that I've been meaning to but uh, I've been buried in um Kind of uh, basically a release that I need to get done for this week Uh, any anybody done anything with aries agent test harness or aries ascar Uh stuff Most of that should be good. Uh, I did post um in our chat channel um the Uh, I uh timo had posted about how they're hoping to handle Connections in afj, which is something that we've talked about with akapai and hakan and judas work Uh, so I did post Sorry, I think I know how to use the internet. Here we go. Um Yeah, basically timo gave some suggestions about their connection record and what they'd like to do Um, I think hakan took a look at it. Did you want to say anything about it hakan? Um, nothing particularly it it it looked like it is very similar to didx change But I think timo replied back to that comment and said it was related to didcon v1 if I remember correctly so, uh Yeah, I mean overall in general it looks very similar to what Endpoint of didx change would do like look at the public did if you find that resolve it and uh Yeah, establish a connection basically Okay Yeah, very good. Thank you Um And then uh, yeah for today the topics that I've posted, uh, so nick reynolds Posted his libpdp Didcon contribution as a pull request Might be worth looking at I started to look at it, but I haven't made it completely through yet Uh, essentially Yeah, he gives quite a bit of information here Um, did anybody else uh take a look at this yet? It's it's nice because um, yeah, it's libpdp has um quite a bit of Functionality added to it including there's people who have built like bluetooth, uh uh transport Support and things like that Anybody else uh know more about it? Well Something to be aware of so I put it up there Um, yeah, I'm looking forward to to going through that Uh, and then I thought for today we could essentially keep working, uh on our Our ari's interrupt profile three, but let's go around and do uh updates first if anybody has updates from their projects And how it's going Maybe uh bruce if you want to start Yeah, nothing's happening now our students are taking their Their went their holiday break I I do Have some questions about Well, just one question, which is Did we get them started correctly on didcon v2? We have them Um implement a did exchange protocol But I believe did exchanges for v1, so I'm not quite sure what you should have them start Yeah, that that seems to be a Repeating confusion between didcon v1 and v2. It's not um I guess it's not intuitive to think that there is not a did exchange step That that that was part of maybe the simplification of uh didcom in didcon v2 um I'm glad Yeah, so essentially receiving a didcon v2 um a message is essentially how you're establishing the connection Um, did we want is is there a good um guide? um or That the the guy would yeah, yeah, yeah Which is good. I think the Hello war Yeah starting using and I guess the thing is does but this probably doesn't warn someone that Things have changed between didcom v1 and v2 For instance, they would be natural. It's not related to to v1. So nothing. Yeah v1 Yeah, so may that's actually fine for the students who haven't encountered v1 It's just challenging for those of us who are guiding them Yeah Fair enough that will help um Yeah, I think uh I guess it Yeah, if you follow the starting and using Starting using and ending a didcom connection will guide you how to establish Normal connections just a simple message. I think that's the way Can be handled easily without the change Of course that's dependent on the library because if uh as hack and point out if you have a library v1 And it's being trans translated to v2 Then that library may require something like like a protocol to To establish the final connection and set up in the in the database But didn't come be too simple stated can be too. It's just one message One way and then you can replay back and establish the connection on the other side Thank you. Thank you, bro. That's that's helpful. I will study this And um the students are actually if I understand it correctly building their own Didcom v2 library, so Right, right, but they do speak python. So I think I think well python. That's python. So So they're gonna They can use it. This is a notebook. So they can test it on On the cloud. So Okay, thank you Yeah fair enough Okay, thanks bruce Um, let's see How can you want to give any updates for this week? um I don't think I have any points from for for this week Okay, there was no additional Yeah, topics that I worked on related to this computer, but maybe Judith has something to um, yeah that And not really unfortunate because I was sick during the whole week, right? Yeah Hmm. Yeah, I'm sorry about that. Yeah, that's that's A lot of people out sick right now and then of course holiday stuff. So Yeah, fair enough. Um, I do think did we talk about it before I guess uh, our We won't meet uh next week Uh, I think so I think I updated the calendar Uh for us we'll meet again. Let's see here Just double checking It could be january 9th from yeah the 9th The 9th should be back. Yeah, very good. And I just want to make sure I removed it Yeah, I did remove it from the calendar. So they gave me permission for that. That's right. So Good stuff, okay And then uh, alex and roto either either review of updates you want to give Nothing on my side Okay Okay Great, and I'm not missing anybody. Yeah. Okay. Good. All right. So let's take a look. Um At the areas in our profile three and Yeah, we could add some things to this Uh, just go through it Uh, I think most of you have seen it, but um, I don't think we've actually just you know taken a session here To go through it If that's good with everybody Um, okay, so Is everyone happy with this initial Um base requirements For ap3 what do we need to add we have the spec did come v2 which includes these protocols On top I think it is interesting that basic messaging is separate Am I right? I think yeah, it's down here Is that do we feel like that's right to be separated from The base requirements Is from my point of view it makes sense because Once we have the did come covered We want to use it for different purposes and basic messaging is one of those Purposes basically I would compare it to having this protocol similar to credential exchange protocols But yeah, on the other hand, we do have base requirements for credential exchange. Hmm It it does seem a little weird to me that uh, that there could be I did come implementation without basic messaging Uh, and then yeah, you're right that then we're showing Transfer of credentials I would imagine Well, I guess I guess it doesn't matter to say which one's more important or or more common, but uh, yeah Any any thoughts on that? Yeah, well, but basically just for a chat application at least this Is for two persons talking right It's just a specific type of message Yeah, I don't only I think it's only can be used for for a chat, right? Yeah Anyway so Did then did we want to peel this portion out in terms of Being able the the wacky well it says wacky packs, but uh Wacky did come Is that something we want to separate as well or or at least comment that that we think should or do do we want to leave it here? I would I mean If you look into the first IEP AIP one and two, I think Those are really related to Aries and that's really related to issue holder verifier issuing credentials and presenting credentials Like as what what what's it's supposed to do? And if you look into the AIP 3.0 also from the same perspective and that's the if Then I think this is fine the way it is because we're covering everything related to ditcom support functionalities to issue be able to issue and present credentials Yeah, I agree. I agree. Yeah Si okay perspective because we're asking our students to Implement I guess to follow the spec for ditcom v2. I guess and We don't intend to issue credentials or require groups We'll we'll be defining a A custom pico specific protocol on top of ditcom v2 and that's as far as we're going to go So from that perspective, it would be nice if if If our students could say, yeah, we're AIP 3.0 with no options Whereas yeah If there's a credential exchange and proof Portion that's required to be AIP 3. Then they won't reach that But that's a that's a strange way to think about it. Yeah, that makes sense a good point Yeah Yeah, for sure And it kind of gets back to what we talked about before where We're trying to help people to know the common path, but also like the The the the simplest Uh support that they need first in order to kind of add the optional things And so we've even talked about wacky being something like oh, you know wacky did come simple You know not even implementing everything from from wacky did come so I do I I am concerned that Someone new coming in they don't really know uh, and then they see this And they're saying okay Like this is what I need to support and then they get into like wacky did come let's say and then they're like Wow, this is a a lot of work. Maybe I won't do this, you know, uh, this is these are the lessons of jff Which by the way, I met with Sharon from jff this week and just was trying to convey to her that It was really valuable uh that they kind of created to the jff challenge to have these multiple tracks with oidc and and and did come now they are focused on transferring credentials, but because they you know created a did come track and and um Alex worked with uh snore and brian rickter and which Maybe alex if you want to give any updates on that you can in a second, but um I guess the lessons learned from that is uh part of the reason that this group group meets is because of of the Difficulty that that those groups went through as they were coming up to speed uh on did come and specifically did come v2 And so yeah, I that's why this concerns me the way that it is Those are my thoughts alex any any updates on jff for thoughts on those things Yeah, I mean, I think what you're saying like makes sense like also in terms of like the watch attack specification Like in the afj meetings, you're mentioning how like we need to upgrade to like 3.1 and 3.2 to like issue multiple credentials at the same time So from that perspective, I think that the issuance and exchanging of credential is only going to get like more complicated in terms of like Uh, I think you can break it down to like hey, you can send the credential Or like hey, you can request a credential before sending one or like hey You can request multiple before sending one or I can request multiple before sending multiple That's like the different configuration of walkie packs per se and uh, yeah, I think It's relatively high degree entry, but I hope that most developers won't have to write the low level code of doing the protocol exchange uh And they just have to worry about like the business logic of like what credentials they're sending through um, at least that's how I'm visioning and an agent on my behalf like I'm not trying to um If I were to be a developer to use an agent, I wouldn't go ahead and like implement my own protocol I would just kind of like whatever that agent protocol has I would just work with that um Yeah, believe my thoughts but just um See I guess one of the things when I was talking to Sharon and She had done a she did a presentation actually at diff the the diff I think interop Meeting, uh, which is what sparked me to want to meet with her. Um but the I My sense is that Everyone came out of this or at least Sharon and people who who are watching closely they came out of it thinking. Okay did come is complicated and so they didn't make as much progress as The oh idc. What do they call it dc or something? Uh track and what was the other track? Was it? It was a vc api like the hcdp Okay, chappie. Yeah Yeah, okay So so my yeah if if everyone comes out of that and and what I tried to explain in the diff interop meeting is Okay, well part of the reason that that did come, uh, you know, maybe went slower is For one thing we're in the middle of we're in this, you know space of you know, did come v1 implementations existing and did come v2 implementations being very new and You know teams just having to decide are we going to do v1 or v2 and then you know a v2 Team kind of formed and people are building their own implementations because they want to learn But then, you know, there's all these pain points about crypto and even just did methods and things like that So yeah, yeah, I agree that did come is more complicated. I wouldn't say it's necessarily like more complicated I think that the problem there is that The other two groups had like more tooling available to the developers. So like right imagine if like Chappie had a problem between like hcdp1 and like hcdp2, right? Yeah, but everybody knows hcdp like which protocol to use So that's kind of like the struggles that we were encountering with our implementation So like a lot of like the low level stuff was not matching up whether we'd like chappie and oidc connect They already had the libraries kind of like working together and all they had to do is implement the flow Whether the the problems that we had was like slightly different than those So yeah, that's why I think like From like a did come v2 perspective and also like we kind of like started late So trying to troubleshoot all those low level A problem we didn't like anticipate for that Yeah, so yeah, that's kind of like The issue there because yeah, like brian was able to Talk to brian like recently and like he was able to work like the the chappie The extension in like two days since like everything was already there and you can then watch happy pretty much But just simply copying and switching a couple of variables around and that would be it. So yeah So is is anyone else in this group kind of concerned about those things or Is this something that you know, we're just we have a recency bias and you know our own experience type bias Should it be our focus As part of the aip 3.0 To make it very clear kind of what is Most common and what's available for for developers as tooling and for us to focus on you know, making sure that that The tooling that's needed is a high high priority and that the community areas community and beyond has it Um How much should those kinds of thoughts be affecting aip 3.0 3.0 Yeah, I personally think we should try to pretty much figure out like the minimum basic requirements Right because right now if you use did come be to you're most likely using the peer And not using any other did method Uh, and that's because of the library support, right? It's not because people don't want to use other did methods And whenever you try to integrate different did methods, it's relatively challenging for the outcome libraries So yeah, trying to figure out like a minimum set of requirements like signature schemes elliptic curves that we're using for like Uh, an agent just so they can talk to another agent It's uh, I think would be helpful. Yeah other thoughts Yeah, I agree. It's just like I think Well, the aip shouldn't matter on the what toolings or or maybe Work on the tooling side but working on putting what is it this minimum and The minimum protocols and maybe even go farther and say on this protocol. We're using that way For example, uh static prd id That's three type of prd id. I go zero one and two I think So so say which one and that way or also maybe on the mediator How we put the routine d id in the service endpoint just to agree and How we're gonna do it for aip three and if you do it that way Your your case will pass, right? So that's So is that how how would you like to document those things? Here Some says for example, for example, this feature 6 27 static prd id Some current says he's gonna update that to he's gonna create a new one with that in mind So he's gonna change that that was uh mentioned the last eighties call last week Okay, probably there we should put like paragraph that if you're using deep peers with a mediator you should put the service endpoint in this this way Right, so everybody can understand When they receive a routine d id on that path Work can be structured Let's see Can you say that again using did peers in a mediator? Yeah, when you when you use Listen, here we were agreeing where we are gonna do did it come be to with deep peers, right? Yeah, the minimum And then if you use a mediator The mediator give you a routine d id That routine d id will be part of your service in your d id document. So We should agree in a way where we should put that d id that routine d id But because the the specification is not that clear There's a field called routine keys. There's a ui. That's something that I There's a threat on the on on the discord uh chat regarding that and I I say something then Steven say another option team or another one brian Different one. So there's many many ways to do the same and the end is the same Yeah, using the routine d id Yeah, okay so, um Any other thoughts in terms of breaking out? Well, wacky did come or wacky pecs as it used to be called Breaking it out into a separate section I would like to put in another plug for For following the layering of trust over ip Yeah See did con v2 On layer two if I follow the spec and perhaps the static peer dids. Yeah creators And i'm a ip 3 at layer of three of Of trust over ip if I in addition have at least one I implement at least one method for exchanging credentials Yeah Um, I I think I would argue that a ip 3 is actually having components or rfc's from all three Bottom layers like on the bottom we have for example the static peer dids Is one of the features that are supposed to be implemented on the area of agent layer. We do have the did come um as an encryption envelope and Yeah on upper layers or like on the Layer three we would have the exchange protocols issue credential present proof That's a good point. I had forgotten layer one which is also involved Because in our project we're only using peer gates We're not using any dvr's right yeah so Lance I I think my question would be We are going to have a profile and this profile is important. So once two Agents are creating a quite establishing a connection using this conv2 like the um, yeah protocols And they would be able to So what exactly what what would that mean like is it? Does it mean that they can exchange a set of Credentials Does that mean that they can exchange messages like by a basic messaging? Or does it mean something else because with the base requirements what I'm having a bit of a problem is Okay, let's say I support all that base requirements, but I don't support anything else I won't be able to communicate. I won't be able to write messages I won't be able to issue credentials because I don't support for example jc ld attachment or anum cred attachment Or I don't know what the third attachment was so Just with the basic requirement. It doesn't mean that I can't even test whether I have a Well, maybe we could test with discovery features discover futures, but Like I said, you wouldn't be able to exchange any Basic messaging or you couldn't be able to do Issue credentials or present proof because of the attachment formats and I think Yeah, but I think the attachments are on the optionals right and then I think you should pick one optional At least the main plus one of the optionals and find another Asian that that has at least the same Yeah, but what do those optionals mean? That's the part that I don't get like is it optional in terms of it can be Nothing it can be you you have to choose at least one or All of them That's that's true. Yeah, that need to mean more Specified they are what what it means, but I I mean for my point of view I would say, okay, I got a I built a nation that has the base requirement and understand shays on ld So let me pass at least all the tests that That made that Correct, right and maybe I need to find an acapiation that has This same requirements plus shays on ld and all those tests are gonna pass So I'm gonna be confident that that my asian that I built Can't communicate with with others in that specifics Yeah boys, right But maybe maybe that that's something in the aip protocol. We should really state that right make a huge part of our staying Okay, this is the way we think This is this is it Yeah Yeah, I didn't think like personally I don't think that agents should like determine what credential formats they use like because do you try to Figure out credential format to use like in the actual walkie-packs protocol like hey I'm trying to issue you like a jwt credential like do you accept it and you can like hey, I cannot support jwt credentials uh Like I'm not sure that you can like do that at the discovery level of the agent It's like maybe you can do like hey this agent supports just an ld credentials only But yeah, like I'm seeing like a multi format world or like you're gonna have to support all of them eventually So yeah, I think it's going to be a matter like on how you want to go about Telling people like hey, I'm only supporting this and then we're gonna support other ones in the future That's that's a very interesting point. Alex you use the word eventually because our Our methodology with picos is that when we can grab some students we have them do a small piece And then next year we will build on top of it in some way I was also intrigued by An analogy Do that will only be understood by the by by people in this call Which is we we were visited What we're using zoom as a transport layer to communicate And we were visited by dimitri Who was able to use english, which is The basic interrupt profile for this meeting And we were able to have a brief interaction with him to discover that we had nothing further to do with each other And that to me that would be the base layer of a ip3 You can you can you can meet another agent on whatever transport They you and they are You can have a brief conversation and you may realize that there is nothing further you can do with that agent and it's and it's over Or you may find as as one of you said earlier you may find that there is some Some further thing that you have in common. Maybe you both do Basic message and you can chat Or or maybe you both have one of the configurations for Exchanging credentials in common and then you can work together Yeah, that's that's really good bruce um, I so Just building on top of that then would we say that um You know trust ping is not necessary for instance, uh, but discover features seems to be Like this initial interop Uh spot and we've talked about, uh, you know, maybe even having something above that that That that people can agree on just so that they can complete Uh the discover features um You know transaction or exchange, um but Assuming that, uh, you know, the other things are taken care of Uh, then do we agree that that maybe discover features is where we would want Um, a didcom Aries interop profile 3.0 compliant agent to start I I don't I would I would say the full spec did come be to a spec Yeah, right. Trust me is well. Yeah, why if you do discover feature what trust me is more basic than that so Yeah, but uh, if if you had two agents and again, we're talking about an like interop profile here. So I What I keep seeing what what I think is a reoccurring theme is that uh You know, let's take bruce's group, right? They're going out to you know, let's get started with didcom And okay, we're going to do didcom v2 and then they have their They have their requirements, right? They have the things that they're interested in and and and you know I was I I showed my uh bias this morning today Basically saying like uh, shouldn't basic message be here because I'm thinking basic message Is something that that you know is this base level thing, but it's not right? I think uh, I think we're basically we're saying that uh, it isn't um, so If if we were going to have someone do I didcom Uh challenge or implementation Uh Or usage I should say and and it's going to be an aip 3.0 Uh Usage then I don't think that trust ping is something that you would it's a nice place to start You know, it's it's kind of a hello world Right, but it it's not the thing that uh, you need in order to be aip 3.0 in order to be aip 3.0 You kind of need to know What what can you do? I I you know, I'm here. What can I do and what can you do? Okay, but for that you also need out of one, right? You need to start the conversation So you need out of one and you need discover feature Yes, I I agree. I think we're discovering the number zero Right Yes, what what is right the number zero and yeah, I would be fine with Saying that out of band has to be supported. Uh, you know, uh, and then so trust being so so simple So simple that that would all together is No, I I agree that it is the but again if you wanted someone to do, you know, be minimally aip 3.0 Do you need trust ping trust pings easy? Yeah, why not if something that we need to to define that's you define what it is. Why discover feature? What it is? discover feature Well discover feature. Can can can you do it? I can do everything without discover features, right? You you feel like you can do interop without discover features? Yeah Yeah, of course. Yeah, it's on the message. Don't give a reply. So that's it If I ask for a credential, we see the credential and we're okay So that's why we're putting all together. So yeah, that's Minimum trust being out of one discover feature Routing if you need it, but if not, it's okay Yeah, the Sorry, roto if if I come to the table saying I'm aip 3 And the other agent sends me a discover features request and I don't respond then clearly We have nothing further to say to each other Which is fine But we interoperate at the zero level In other words, we can hear each other's messages And and we can then determine that there's nothing further to do But that that does feel like You're doing did come And that's not quite like a interop profile So how come you've got your hand up and how can I had said like, you know, what is the point of of aip 3.0? I'm trying to You know try to add different constraints on you guys and see what the what the response is In order to kind of Fair it out like, you know, we're not talking did come v2 here. We're talking aip 3.0, which uh, you know needs did come. So the question is like, but what is that minimally? Look like so that we don't put any extra burden on on an implementer Again, if they hook up, let's say their agent to the ari's agent test harness and you You know, literally have a tag in there that says aip 3.0 Which means run run these tests run my agent against these tests that are aip 3.0. I would love That to be like, hey you achieved we'll say level zero of aip 3.0 But the question is what is that, you know, what belongs in there? So that's kind of the way I look at it How can yeah um I think I agree with you and also According to these base requirements We could actually detach the issue credential and present proof from from the base requirement itself because If you look into The other requirements that are here expect except the revocation notification now Um, it's it's all related to did come so that you can actually establish a connection using did come via out of band or Trusting is a part of the did come spec. You discover features the part of the did come spec routing is a future of the did come spec so if we want to do the did come be to If you want aip 3.0 based on did come be to then we would need to support these Protocols that are written within the spec of did come according to the diff so that's why I think we should have all trust pink out of band discover features routing But this is something that I don't think we should Dispute because it's already in the spec and if you want to dispute it Then I think we should bring it out to the diff working group um and also Yeah, my second point would be if we want then we could take the Exchange protocols out of the base requirements and put it into credential exchange area Yeah, I support it. Yeah, because it's we can write the Aip 3 in a different way as the other and say exactly what what it means Little zero or whatever we want to call it It just did the come be to We did it peers and then the others. Yeah, like credential exchange mediators and Yeah other thoughts. I mean I I Are people warming up to this idea of the aip 3.0 having kind of a level zero and maybe even a Maybe we have other tracks that are like uh We'll say like credential exchange level one or something like that like a way of of again guiding somebody towards like here our minimal common paths that will help you to Hopefully interrupt with someone else for the the use case and the requirements that you have Yeah, it's to me level zero sounds like the big conv2 spec Uh in terms of like the features that like I'm hearing about is like, hey Like you they implement the conv2 spec or I kind of like you don't um and then uh Yeah, so like I feel like that We can already have that like with the peer libraries and You can do some like you can kind of spin up with some minimal scripts in python, right with the the sick by libraries we did peer so you can have like that Demo like interoperability But once you really start to do like anything that's not Demo like It's going to get challenging to figure out exactly the credential format signature types Making sure like everything works, right? Like all the models are mapped properly That's when I kind of it becomes like more dicey I think they will already have that level zero interoperability level where we can talk Between different like transport layers since that's defined the big conv2 spec in it of itself. It's transport agnostic uh Yeah, that's kind of like I agree with arcane as well From from the perspective in terms of like we should implement at least minimally everything that's implemented in the spec okay now is that In reality is that is is that what's in the sick by libraries right now for instance Yeah, because like the problem that we had with jff is that snore and brian they used the did web So had they used it peer it would have been like A breeze So that was the only issue there. It was like a new method Problem like it wasn't like the the messages themselves were like too complicated or whatnot The reason why we had to break down the messages is because we didn't have enough time and The encryption wasn't working. So like we tried to like reduce The message exchange so we didn't have to implement all the underlying cryptography since it wasn't working in the first place But had we had all of the low level stuff implemented like Writing walkie packs as an exchange. It wouldn't be that big of an issue because If you just support jwt all you have to implement is like those fose library and most developers are like familiar with jwt's and and whatnot Rather than jsonlb or an on creds or The iso mdl spec that eventually is going to be and whatnot okay so Shall I comment then on on pulling these out into as optional Is that is everybody okay with that for wacky pecs for for issue credential and present proof And I think relocation like because relocation is regarding Okay, so all three of these Uh suggesting Is everybody okay with me putting the comment that the aries did conv2 working group Would like to consider pulling Revocation notification issue credential And present proof into An optional section. I guess just to be able to advocate Aren't we just then implementing that it can't be to spec if you don't have like the issue credential in there Well, I think that the at least one difference would be specifying the peer did And and and that at least shows like if we were again to put it in in terms of the trust over ip stack um That would show layer one, you know, what is you know, it's nice to have super simple layer one You know go peer did right? Uh layer two Did come v2 implementation, right? And then essentially we're saying layer three stuff is all these optional things That maybe we have uh, you know documentation that says Okay, these things are optional, but you know, here's here's a common path for issuing credentials or you know, whatever it might be Thoughts on that. I mean it seems like very little but it sounds like Alex for jff for instance, you know, it would be what you guys would have been way more successfully if if that had been, you know, specified That way possibly Yeah, definitely. I think it would have been helpful if we can just have agreed on like using deep tiers Uh, yeah, but we didn't know that the libraries support for adding new deep methods were that wonky Yeah, yeah, and that's yeah, that's what we want people to kind of understand is You know, okay, maybe this this may seem trivial to you, but uh, you know, trust us it's not right or something like that Uh, and there's probably something else that we're not quite thinking about yet that might also go in kind of this minimal You know layer one layer two There might be something in there that we're not thinking of although I wouldn't know you got you would know better so um I mind your point Lancer could soften the blow by By uh saying into a recommended optional section or something like that. Okay. Yeah Then that that's a nod to the fact that it that it had been in the base right as some someone said earlier, I think I think it was alexa. How come um, we eventually we would want to support not just The the walkie-packs way of doing credentials, but also at least one of the others and perhaps all of them eventually but But you work your way up yeah and yeah Sorry go ahead Okay, yeah, and Right and so that might also be our opportunity to say uh, okay Here's this thing called wacky pex light or wacky did come, you know, minimal or you know, basically saying Okay, this is inspired by wacky did come If you implement this your your you know, we'll say we'll pretend it's 33 of the way there for for wacky did come but you can start interupping with other people who have implemented You know wacky did come simple Does that seem reasonable alex based on your experience and and or yeah, and anybody else thoughts Where we've got uh four minutes three minutes um I'm trying to think sorry, uh, I guess like what I'm trying I'm not sure that So like the problem is not like the the wacky pex itself I just defined like the messages they need to be sent back and forth And like that's not hard in and of itself, right? Like you just it's some headers So you got to put on in different messages. The hard thing is like doing like supporting the actual description And like the do you support credentials definitions credentials presentations? um, this for json ld Like a json ld processor Um, those are like the other things. It's uh, not necessarily like I don't think that wacky pex like whether like you asked for Getting a credential first or not Is that challenging is actually like doing the validation like do you need to do that? Or all that kind of stuff because like you can always omit it In the wacky pex like some steps are optional already in there So like you don't have to request a credential you can just like pay like propose a request And then like accept the request and then like you can issue a credential and then accept the credential So I think it's relatively minimal already in the sense that it doesn't define how to like send multiple credentials at once Or that kind of stuff Yes, like i'm not sure like Breaking that down even like the wacky pex like even further because then all you just do is just like issue a credential and That would be it In my opinion because yeah, like the credential manifest is already optional in wacky pex Tell me something. Yeah, so um If you When uh, when if you do discover features Does it tell you? Hey, uh, okay. I support uh wacky pex But I didn't implement these optional pieces I'm going to assume it's not we or traditionally hasn't been that expressive not that you couldn't do it Uh do it using discover features, but you get what i'm getting at right that that again this discoverability Okay, if there is a bunch of parts that are optional, uh, how do I know if if you did it or not? Do I just start we just try and then we break or you know, it would be nice to you know Be able to discover ahead of time uh I guess, uh, you'd be nice to discover ahead of time to do Like what's the use case? I guess like i'm trying to figure out. Hey, yeah So I want to do like request a credential, but like you don't support me asking for credential Like what's the I was I was thinking more like uh, so you had mentioned like credential formats For instance, what if what if you wanted to say something like hey, my preference is that you give me a credential in uh json ld But so you do that you do that whenever somebody like in the wacky pex protocol That's when you do that. It's like hey, I'm trying to send you a jwt credential Do you want to accept it and then you can say yes I want to accept it and then like the next message is like somebody that's issuing you that credential With the data that they send you in the propose and then you can accept that or like you can decline the proposal So like that's when you would do that Yeah, fair enough It's like hey, these are like what proof types do you support? What credential do you support? Like I'm proposing this credential Do you want to accept it or not? So that's already done at the wacky pex level like let's get negotiated in in there. Yeah Okay Yeah, gotcha Okay, any last parting thoughts because that's our time um Is everybody feel all right with the direction we're heading Yeah, maybe one thing that I would like to add or ask to the aries community because I mean most of The people participating in here is if I understand correct not using aries framework as as a underlying Yeah framework, let's say I would like to ask this as well to the aries community who are who worked on the aip 1 0.0 and 2.0 to get like a bit of background information about like what was the perspective that they were taking on Like was it the perspective of credential issuance was the perspective of fulfilling the interactions that are happening between issue holder verifier Or was it more general and depending on it. I think you will have a better direction of like having the aries into a profile tree Yeah, good point And learn from that perspective, but then I think bringing this new perspective of hey This is a group of people who are who I actually have to deal with did come be too, right and and and you know Maybe the perspective has Has shifted now, you know, obviously bruce is uh, okay. Yeah. I think that's perfect Okay, well, I think there is a meeting the user group meetings later today for for diff And so that could be an opportunity and then aries working group Uh, has been covering aip as well. I think on wednesday, so that's been really good with sam So, okay. Did you send me that link? Please lens the one for the aries one. I'm not aware of that one. Thank you Yes. Yeah, sure I will do that right now other than that any other parting thoughts Happy holidays to everyone Okay That's right. They'll be celebrating for four years now Yeah, that's good Okay, the yep the link to the latest aries working group people are you know, obviously free to jump off But I just wanted to get that into the chat Very good. Great to see you all Uh, we'll get together in the new year. Yeah. Goodbye. Good. Thank you. Bye. Bye. Bye Thank you