 Hey folks, hello, how's your working from home going? I'm getting used to it including getting a new webcam and Taking all of Scott Hanselman's advice. I can get I'm not sure if it's glow light thing, but we'll see how that works How about you is there anything? Well, I guess you are always working in. No, you're not from holiday. I'm actually I'm actually in the office today, but I'm Yeah So, yeah, I don't know. I'm probably gonna start work. I've been working from home some of the time I haven't gone fully home yet The office might be the better place where nobody's that Well, yeah, there's none of us. It's none of us the sick yet. So I don't know We pull up Here Our hack I've been debating like do you put the background on it's in fact I was gonna ask Phil if he's got the background from my office because he was he was put that on the other day It's interesting to have us all in neutral territory now It's like there's been so many people that work on I worked from home already picking up new a new appreciation I'm just looking to see if Omar is joining I think Sam is on vacation starting today, let's see Slack on here and just for everybody's Reminder we we have the auto-recording on now No, Justin. Did you have any ideas on what how you wanted to start out the break out? Oh, there you are great I think of Omar and he appears It doesn't happen often Mostly I'm like somewhere else We were just discussing who's working from home and who's the office and being at the office might be the empty place these days. Yeah Okay, we got Justin here as well. That's awesome All right, Omar you want to kick it off? Sure I think a couple of things I want to do was first one We should chat about And this may be quick is what do we want to do about meeting? Now that coupon is off So I did Rescheduled to an online meeting I moved it from in the middle of the night at least for the US folks And I think it'll be early morning for the UK folks to be 9 8 9 to 11 a.m. Pacific time so that would be What time is that for you Justin? We just switched to daylight savings. All right, it's really confusing. I it's it's actually 9 is now 5 I think But it but it we're changing time in three weeks. Oh, that's right But actually we're probably I don't know. We're probably changing just immediately before that So the body would be back to sex again. It's so confusing. It's really am Is daylight savings time going to be cancelled also? Not in the UK, but it well, no, but Europe's gonna give it up apparently but the UK is not gonna give it up That'll take three years for you guys to do. All right, I'll stop with the jokes I'm happy to meet at 8 a.m. For you know, but Kurt meetings 5 to 7 p.m. UK. Thanks, Joshua I'm happy to meet at 8 a.m. But I recognize many of developer species may not be too cool with that on the West Coast John is the only person I'm recognizing on the West Coast. I guess Niaz As well, can you guys vote in if you're okay with that? Hey, it works for me Okay, and John would you be willing to get up at 8 if you were joining? Oh John sorry, I was looking there you go. Okay, so I will schedule it for thank you, Justin Tutan, that's right Yes, list is growing quick. So I will move that to well We to be fair two hours was probably good for notary, but we had other conversations. We wanted to have Which was the original plan at Kubecon So I will Let me let me think about how we want to do that because there was notary there was artifacts and there was The pub sub stuff that Joey was working on that we were thinking about discussing so let me see What the list of agenda items are and then I'll propose some times and we'll go with that but I'll start whatever we wind up doing I'll start at 8 a.m. and Because I've got our UK folks here now. I'll start the 1 and 8 a.m. for notary just to give Justin half a shot of Being cool Yeah, I can do that. Excellent. So that's let's just close on that so we can move on Cool. The next thing to chat about I think is Making some progress by dividing and conquering. I feel like we've done a good bit of work on scenarios and getting Chatting and getting enough in the community Getting going so now I think a lot of people know what we're trying to achieve It would be nice. I think to take the next step when we first kicked us off. We spoke about breakout groups right smaller groups getting a little more focused on trying to drive on different paths and I'm looking at the meeting notes the hack MD And we had asked for breakouts, so we had at least discussed breakouts one around threat model Which had key management as part of it One with use cases which has already begun I know Steve and you've been doing good work there with Sam Justin and Justin Kapos, which has become scenarios and one around signature storage Which Has Vincent and Derek's name against it as potential just you know pushing for it Do we have either Vincent or Derek on the bridge? I Um, I don't we don't have Derek. I don't think I was gonna ping. I'm gonna ping him separately Vincent I can also paying I said Sam obviously Sam wrote up some notes. Yep Which we have which is great And I've talked to Derek a few times, but I'm I'll try and Yeah, I'll try and catch up with Vincent see where he is as well and see if we can put something together on that Because yeah, I think we've got some we've got some rough notes and we need to turn them into Things that everyone agrees with Yep, so I think one was threat model key management, right Justin You were you were going to target that and I think from the AWS side at least We had a couple of we had a couple of participants from our key Signing crypto team both Aaron and Nia's so we can most definitely participate there I don't know what else and who else you were thinking well, just Justin Kapos and his team were interested in that as well so Marina as well, I think I saw some notes. Yes, she's on the she's on the call. Um, so yeah, so Yeah, although what one thing we may do is Because there will be different working groups than her and I and Santiago may it's like spread out across working groups and Okay, because they're saying because I think that might be more valuable for everybody Yes, I agree Justin that makes sense There was another working group. We were thinking of around the user experience basically, you know, the API to go for a sign Harder to work in every one, but we want to do the two and I think the scenarios and that kind of go well together because you're going to approach it from that sort of use case view of life So Steve, maybe maybe you and I jump off on that one and Justin pick somebody from your side as well Who wants to participate there? Um, yeah, I probably will take that on my side Okay, so you want to you want to be on both then Justin, yeah, okay, cool Then we've got those two working groups there was One of the two working groups I was typing the others I'll try to keep up with notes for while you're talking to ours in UX like Because the scenarios in the use cases will drive the user experience right the API experience So I think we should think of those together and then the threat model key management as one threat model and key management, okay UX and sorry UX and what I got a speaker setup here use cases in UX, okay And I'm feeling like we can maybe I don't know what do we think should we continue this call and then also have a call for working group Or should we just make this a half hour or just eliminate this call for a couple of months or make it once a month Like how does that work to stay in touch with everybody but yet not duplicate meetings? I think we do need some sync for sure across these because Obviously like UX can't go make a decision independent of like threat model and So on so I don't know the best way to coordinate, but there has to be some some sync meeting Yeah, I mean why don't we Why don't we keep this but keep it short if we don't need to sync very much and Because I think yes for some for the maybe for the next few weeks half an hour might be Good, but then we might if we've got Working groups coming back with detailed proposals that we need people outside the group to review then we might be longer again And we could treat the Monday as a stand-up. We'll give reduce it to a half an hour And we can make sure So this way and then what would be really good is if we could so that not everybody has to attend all the meetings If we wanted to have a designate for each one of those that can be at the Monday meetings Not that others aren't welcome But at least we know that there is one person that is designated to kind of keep as a sink on things and If they can't make it they can coordinate with other people in your team to represent them Yeah, that works that that sounds great a Monday stand-up weekly Yeah, and then each each working group and right now that we're thinking to can Figure out and work their own cadence on on the progress that has to be made and we can just dump everything online as we're doing today with PRs That's good And I'll also reduce it to 30 minutes just to Give people time if they wanted to schedule their own like people want to keep the hour blocked Blocked for themselves and use the other 30 minutes. That'd be great In fact, how do people feel about me making it the second half an hour of this slot this way anybody that wants to Do their meeting they can do with a 30 minutes prior Well, then we can't use the CNC of no to resume at the same time with groups But I mean, I don't care we can happily use Chimers something Yeah, we'll let the leaders of each group kind of coordinate that out. Okay, is there any broad based Groups that we all like aspect of this we are missing that deserves its own breakout group Like we had something on signature storage and OCI Relationship or how it syncs up so Yeah, Justin, didn't you want to do signature storage as a separate because yeah, yeah, yeah, so that was the third one you know Yeah, so that's that's this that's the Vincent Eric Sam I Think that'll be quiet. I mean, I think that we have we have some quiet concrete proposals there. So I think it's a matter of just Spending some time just going through pros and cons and making a recommendation making recommendations and working out what the gaps are Got it Sounds great. Anybody from that named of three on the bridge No, no, I will not be here I will I will I will I will talk to them all and to bitch it together over the next day I say it's a Perfect. Um, well other than I mean Sam's Sam's off now, isn't he but we've got his what he's written. So Justin, can we put you as head of this team? okay, I Mean, I will try and delegate it because they're the people who know this stuff, but I will Yeah, I will re-delegate it to someone else, but you put me for now, okay And by the way people should I'm putting some names next to these just to have Something as a start, but please put your names in yeah, yeah completely Yeah, the them if you're interested in specifically in those areas just add yourself and then I'll make sure that We'll make sure that you're invited to the meetings. I Will also price them all and like channels are fun. Who's not here nice Yep. Hey How just as for open items we were and I know everybody's busy and certainly things gotten crazier recently but Naming that we How do we want to come to closure on that? Because it seems kind of important we start thinking about what exactly we're putting a signature associated with to and storage What do you mean? Sorry, but I mean, I think that that's a combination of Mechanisms are you what we're signing how we're signing it in OCI and then the about The threat model and what we're what we need to sign to be secure And and to some extent use cases of what we do I don't want to sign for But I so I think it's kind of I mean, I think that yeah, we will have to put this all back together into a thing that overall makes sense, but I'm trying to understand what you suggest like are we Thinking like we've talked about whether we just sign the digest the tag and name doesn't matter or yeah I think and tag be the sign thing or yeah, but I think I think we should be directed to some extent by the threat model and and that work because it It's it's not just a kind of And and to some and to some extent UX like what What are we trying to? What are we trying to protect exactly and I think making that precise will have one so that's a question Okay, so you think rounding out the threat model Will make that yeah, I think if we do an iteration round Listen detail it'll help us answer those questions. Okay, because I think we kind of just But I mean we should definitely come back to it in one of the meetings again soon, but We need to find a better way of answering those questions, and I think specifics of things like threat model and And UX as well are important I'm worried about this going a little too far being completely open and we're going to get on various different directions But I might be overthinking that concern Even if we're going at one way and changed our mind to have a reference point. I Definitely agree. We need to have a concrete proposal For that, I mean we want to I mean our kind of roadmap was to try and have a concrete design on the cube content table And I think we should try to stick to that so The original coupon date to be fair, so yeah, the original. Sorry. Yes, the original Yeah, so yeah, I think Let's Yeah, I mean let's in parallel. I think in parallel. I think having some detailed written proposals that we could really Discuss in detail that the Monday meetings would probably be a good idea All right, okay, so Proposal to a working group up in the github that we can all read Is that what you're thinking? Is that how these things go? Well, I mean, I think state is Steve is I kind of asking for an overall proposal as well as working group which I think is Not unreasonable, so it's a trend draft something Okay I'm seeing names getting filled in against these working groups, which is awesome. Yeah, thank you Justin CR is that Cormac? Yes, okay. I just put CR because This is today's I think we could use some Some representation from red hat and a VM where I'd imagine I know Saja was from your side Steve Saja was opinionated on a few things and had some good insight. I wonder where he would He's probably gonna be more on well, we'll have people on both key management and the signature storage So we'll dig in more there I just didn't want to put their names on there yet. I won't let them Because she weighs like she weighs actually in China. So we have to figure out time zone. Yeah Not Joey or I know Josh was on the call or Well, not Josh was sorry Evan, but I don't know Joey's join many of these has he Steve Joey sure. Yeah, I Know but I've been chatting with him on the pub sub stuff. So I kid I'll ping him again on this I was trying to get him involved in the notary stuff. Sorry in the artifact stuff is what I meant I'll leave it to him on to what where he wants to spend more time And I don't know if Jimmy is more interested in this like we'll make sure we ping the appropriate folks there because we recognize not everybody can be at every meeting but They can read the notes, I mean at some extent we want we're hoping people will self-select yes To represent their own interest So then goals next Speaking specifically when we meet at the end of March I think it would be nice to have at least a rough draft readout of whatever each working group has managed to Come up with and how much progress has been made I'm I think without having something we're going to try and aim at and get structured on it I'm not sure we'll be able to keep the cadence we need I can I can commit about a couple of hours a week I think also like 30 minutes for the meeting and then an hour hour and a half of chats a month the working group I'm I can commit back Okay Yeah, I put in the notes that so coupon was the end of March first week of April So we're going to just keep that data in there wrote that in the notes that will keep to our original April Original April timeline for initial design. Yeah Justin as well. Yeah. Yeah Okay, cool In fact, maybe we could use that two hour chunk to kind of use as a forcing function to be The review of where each team is at Let's let's do that Yeah Not to get too ahead of ourselves, but I was thinking about when to be able to work back from a formal we're done ish date and There's coupon in where where's the US coupon this time? Well, Boston Well, we have the You know assuming if Amsterdam doesn't happen. I'm assuming us won't happen So the Amsterdam one is supposed to be moved to July or August And then coup con us 2020 is John says November Boston yes, this is November 17 Boston is it too aggressive to have a proposal out by the July Amsterdam Thing and then like almost a rough code implementation by November or I Know if we don't set a date we won't hit a date. Yes Yes, so And at least at least on the AWS side, this is this is driven to word-solving customer pain Right. So at some point this is going to go from It's you know getting something going to we best start implementing and providing some sort of a concrete plan internally And I'd really love to use notary public loan signing to do that. So Yeah, I think those timetables seem reasonable to me Okay, let's let's use those dates and I mean look we all have pressure like I I'm getting pressure on MCR and you know all these different things that Why can't we use the signatures already there and I have to I spend more time explaining why we can't use them Then making progress and why we're doing the new thing it feels at times. So I think we're all feeling the pressure Let's let's see how we like I'm gonna put I'll sketch in like I think it's a good idea. Let's sketch in the April date being our design review and the November Kubekan date being a Working implementation that people can start using Like in the November one, I almost foresee Getting up on stage and being able to state Look, you know in July a Proposal was issued and since then we've been using it and playing with it and tweaking it and as of today We can state that we have an implementation going where it doesn't have to be live does not be GA but at least it's something we have Worked on inside of AWS. We've put engineering resources towards a view to implement So that's kind of what I was thinking Okay That's all I had Sounds good. Anybody else have anything? We have a bunch of people on the call. Yeah, what do we feel about these dates? others on the call Just personally, I think it's a bit aggressive, but I think it's good. It's good I think it's totally doable the notary we won was done. I believe in three months by a few people a docker So it's doable That's what happens when you have highly modded individuals and a small team We'll pretty much go up and do it like that I know We'll just go into self-isolation for three months And dated or not one way or another Okay, I think Broad-based stuff. That's all I had and we can use the rest of the time to to talk about open items or we can use we can give everybody a half hour back and We use slack I guess right stable. I'll take a look at these notes. We can use slack to figure out how each working group Yeah, we can split out more working groups of people want there's nothing it says that the notary v2 has to be a single-player Although to this we'll leave it to the teams, you know it Sometimes it helps. I don't know. I don't want to make multiple channels with no conversation Yeah We should self-select what they think is the best thing But I would encourage people to keep the conversations on slack and so we have at least one place for people to look Yeah Marina you're on slack, right? Yeah I did have the PR is that I need to take a look back I thought I just did a fix quick fix on a read me and I'm not somehow I messed it up or something I got to look at what Sam had commented on it, but I did put the DCO stuff and Excuse me the DCO stuff and I think I got the two approvers Setups you have to have two approvers to get a PR merged Just to have some assemble in some structure not that we want to be overly formatted where things sit sitting PR review for 16 months, but Just a quick act for exactly the reasons that I was I apparently pushed something that Seems to have an error, so I'll take a look at it Okay, yeah, sorry. You actually is that what do we say about the July date? We said March 30th is July our design No, no, I think July could be our first notary Spec proposal or proposal spec or whatever like okay, but not a not March 30th But March 30th March 30 it would be the first time we can all sync up and see how the groups have made progress But I think if July Amsterdam goes on we should be able to go to da, you know Here's what we're thinking and then have ever been throw things at it cool Justin is Derek have bandwidth. Is that what you guys are kind of I'm just I've just pinged him I will make sure I think hopefully he should I mean he's I'm sure he's gonna bandwidth to at least Set down his opinions I can because we've already we've already had we had a bunch of discussions about this before It's just we haven't got it written down. So But yeah, he's got a bunch of opinions about the things that we're talking about about just performance of not doing multiple look apps for things and those kind and Various other pieces of just how how storage should work in and the registry that I'd like to have written down I think the other thing we have to Start thinking about is who's actually going to do the work like who I mean how much of notary public loan Can we just lift and shift over and just tweak some of them? You know the naming stuff that we do you know the namespace and registry stuff But where is the actual code change and repo not just where the repos but Who's actually going to be making commitments of time not just commitments of code? Yeah, I think and there's also some questions about Like because there are a lot of different registry implementations, how much are we going to? Be able to share and how much are we going to like have the reference a reference implementation for You know the open source registry and how much are we going to? What kind of libraries do we want to be able to reuse as much as possible without wasting our own time? That's a good point. Can we how do we want to like? You know we we think of these as specs and that's obviously the first place we should start But I do I had been just assuming and I'm actually intentionally using the word assume here that we would have a reference implementation and I think we should we definitely want to reference implementation because Yeah, otherwise, it's just I mean because we at least want to reference implementation with You know with a docker distribution Registry spec because that's where all a lot, you know, it's the reference implementation of most of the ACI staff and it's where a Lot of people use it as their implementation in some form might be modified or not And you know, and then we need and obviously that we need stuff that goes in, you know for cube and things so So we need, you know, what do I? You know, how do I run this thing in cube? How do I run this thing on? Container docker CLI whatever it might be as well, which is obviously kind of be Like libraries or referencing limitations or Okay, that's upstreamed or whatever it might also a mix of those things Probably a lot of shared library mutation Yeah, so so depending I mean I think it's a little early to see how this all Kind of pans out but Assuming that, you know, one possible future is is that if notary v2 looks basically a lot like what tough is then One thing that we've done for instance in the automotive space and other spaces is that we have an overall spec That gives you all the security properties and things like this and has sort of the security proof and validation around it But then individual implementers write a document that basically explains How they're doing very specific things around wire line format And things that are more like kind of domain specific to their deployment so in It's not um, I think unlikely that this effort would end up in a position where Effectively you already get to inherit the tough spec as things that you follow And then um, you know what you write is more what I think I don't know if marina has talked about poof stuff on this but it's a document we call a poof which is a protocols operations usage in format which actually goes through and um Describes like kind of the wire line format in deployment in a little more detail So that might be a way to um You know get somewhere quite quickly with this Uh without having to redo an enormous amount of work or copy paste things out of existing specs and then I'll Like evolve independently in weird ways and so on It brings up a really good question of how do we think about notary v to And registries like I and I'm biased. I've been thinking about those of being synonymous um Is that Well, just is that what how do we think about this is success that? We have a notary implementation with the registry Or is success that notary v2 is something that stands alone and can be used by the registries Well notary itself can be specific to registries if you want but that doesn't mean that It that the tough like the tough spec if it obeys the tough spec then You've sort of made opinionated decisions to make it work in this part of the space Which is fine. That's that's exactly the purpose of having this like proof document there So you can make opinionated decisions. So I I don't see the two as um Like an either or thing that has to be decided now I think it's it's like you could um Yeah, like like um it it's it's not like it has to be one or the other His talk could be that thing In these strongest semantics you want and to end verifications, so you must have some notary component at the place where the software is created and somewhere at the uh place where the software is consumed Yeah, absolutely. Yeah, pretty much the opposite places of where the registries are On the other hand to the extent that the registries are involved in transferring the information. It's much better if it's uh Included in the registry server than having to maintain and mirror separate services So, yeah, all of them that it would definitely not be the same thing as a registry But that there pretty much has to be some integrated support Correct. I think that statement of what is the definition of success for us? That's that's an important one right to what um I just heard that makes sense. You're going to sign it somewhere You're going to push it off to a registry and then you're going to pull it from there on to a runtime verify that it can be consumed And it's it's it's an authenticated artifact that you pulled that and to end that success, right and Yeah, we have to be able to show that and Ideally you move that object from one registry to the other and do the same thing and then you've achieved your portability goal Yes, yeah, and you're done right at least that you hit that minimum bar. I think I think that's the way that uh, Justin Kappos was kind of referring to tough in the automotive space that made me ask the question um, because I hadn't thought of tough being um tough and notary being coupled in in part of what we were doing but We do and I I wrote the next sentence that we the ability to pull an artifact from a registry and validate it outside of a registry Is part of our success. It's in the initial scenarios um, if something lives completely out of The registry and it's not even you know because artifacts or anything Is that really what we're trying to scope here? I'm not saying that should somebody shouldn't I'm just saying what is what do we feel is the scope of what we want to accomplish? I mean, I can only speak for the tough project, but we are absolutely trying to support all those use cases And we're with the communities. We're working with that have a domain specific use case Like, you know, for instance automotive then one cool thing that we're able to do is we're you know, they go and they say Hey, we're encountering this You know a crazy situation that we've never heard of and it's like, oh, that's really a lot like the situation that we've We've seen from this team of lawyers that's using tough to Manage like provenance information for documents or it's like this thing that we've seen that we know the folks that docker are doing So, um, we're kind of able to in some cases pull in Things from a broader community, which I think has been helpful Hmm I make sense, but Steve to your point I I think that if I think for us at least I'm looking at notary v2 In this group as a function of how it interacts in containers Right Yeah, justice. I think sorry go ahead no more signing an artifact into a registry pulling it down in the client runtime making sure it's Yeah, I think your your non registry model Is not I mean, I think we're intending to support it by means of having A model where you get the same information From a file system as a registry in the same Format essentially in the same formats um I'm making sure that works So that people who are offline or whatever support that but it's um Um, but we are assuming that there'll be a just a mapping between what you would have in the registry and the file system So that we're not building a separate file system specification for how to store images that's different from what um What the register how it would be if it was in the registry so it's a kind of It's a special case rather than the at a different case that Would need Something different so the the validation from the client side and the the way you create it would effectively be the same Yeah, and I would yeah Go ahead As john said on the chat Yeah, the sca image layout is the serialization of the registry format. So Essentially it should work with either Yeah, the thing well actually where do you go ahead, you know, you don't speak And I it would be nice for no review to not to actively block the scenario Of signing artifacts that are distributed outside of registries Yeah, it's not so much that like just two thing that actually brings up a different part So we've always said that regardless of another artifact and how it signs itself We wouldn't you know if they wanted to wrap it in a second signature that moves with it across registries That has some additional meaning We're not suggesting it replaces anything and you know singularity being something already has a format and seen that being one that I know you guys are evolving one rapidly So that's that's something we would want to maintain regardless the the thing that I've been Watching and for Justin Kappos and you know with the stuff that's going on with Santiago and the sbom stuff Is I've been trying to hold a a relatively tight line or I have an opinion of holding a tight line That this group is really focused on the signature of a content and the It's a minimalistic of metadata of providence and all that other detail like those are awesome things But I kind of think of those as this additional sbom document format that others are evolving And how that comes to fruition and so forth is great like at the end of the day We should be able to sign an sbom But all of the details that go into the sbom and everything in my opinion is not something this group is trying to Define we want to support. Yeah, like completely it completely I think we need to be yeah, we we need to have a clear escaping of what Is in and out and how how it relates to these other things As part of the specification so that it's clear what we are and I'm doing and what What exactly the you're getting And I this is part of what's in the root of the requirements document. There's that picture that wait where to go Scenarios, let me just I'll paste this link in here Where I tried to scope that and there's a color coded picture in there that kind of Identifies the end-end scenario including the sboms and all the stuff that tough and in todo and other things that will Produce information including the stuff that you know, Vincent's been working on with the build Source to build whatever that is That all of those things are things that we would sign with a little blue signature And we would support these end-to-end scenarios because honestly without these end-to-end scenarios We while we did was give an interesting signature that has no meaning But but I don't know if we want to conflict we would contribute to and we should you know I'm hoping people split out of this like I know Justin's team is working with the sbom the 3ts bombs stuff Like I'm hoping everybody from this team or others that are working on these various efforts will collaborate here But like and I pick on red hat being one I'm not sure if it's accurate or not but I think red hat has its own sbom format that they've been using and whether true or not I'm sure there's some other company that does and we want to make sure that we support them too Does that feel in conflict with anybody's Assumptive goals we're doing here I guess just in capos that's probably more towards your team, I guess I mean, I think in in general what you're saying makes sense to some degree, but I also think that until we get to the point that we understand like the threat model and things And even like the ux like our goals there I feel like it's it's a little early for us to try to exactly define Like a scope that we're absolutely sticking to because You know as we all know like security is often Lost in the gaps between things And I think right now trying to say we will go no further than this step in exactly this way is um Something that's going to come to bite us in the butt So I I want to just urge to keep a mildly open mind without it being so open that our brains explode because there's We're trying to do everything in the world yeah I think the balance is absolutely the right point And including like we also know that we've got some really secure systems out there that nobody uses because they're so hard to use They just shut them off. So I think to your point like I don't know how far we'll get on the ux I think we'll probably have some rough sketches just to kind of set some expectations And only after You know, we've got further in the design and we know what we're naming and we're you know That will and we don't even know what clients were working with you know, that will have some better ux on it But I think having a I think if it is you're building a house you're sketching something There's lots of details that will evolve along the way But at least we know if we're building a house on the beach or house in the mountains or I don't know bedrooms bathrooms blah blah blah Cool I actually have a hard stop unfortunately. I'm running five minutes over so I'll I'll drop but we still got some time I think I'll ping on slack at least for the first working group and we'll try and see if we can get calls going Sounds great. Yeah, I just I just ping dark and he's got some time to work on this I'll show you something Great. All right, so I'll encourage people to put their names on these notes So that we know who to contact for What group should be, you know, who to keep in touch with who so they know as they're building these groups Who's interested in them? So thanks folks. We'll see you next week. Stay safe Yeah