Summary
createdVia PR in draft pending broader CGS architecture discussionThe following features were discussed as medium-to-high priority for the CGS/EPDS roadmap :
Key technical context :
destroy only orphans the group on the PDS — data may still appear in the indexerall.hypercerts, app.certified, hyperboards) which have been defined but not yet published to the protocreatedVia PR — not needed for current use case; keep in draft pending broader architecture discussionNotes
Transcript
But You have a celebration. We have many celebrations today. But first celebration is... Evelyn's, for Eve's first day, officially, Yes, welcome.
Wonderful. And also, I think I've mentioned it, but just to also have the transparency, Evelyn is with 70% for time joining us. I mean, as we kind of all have some other things going on as well. And for Evelyn, it's like 70%, which, yeah, I'm very excited about. Um, We have more celebrations. And Shafi was on the call earlier. The My Earth Round is life. And we got in the first hour, 5,000 USD donations, even though we didn't communicate it really much except that.
I mean, so this is like also like, so a hundred, no, 200 projects will crowdfund together. together, right? This is like also like why this is actually like kind of is potentially able to like really create more than just one project doing crowdfunding and trying to get the word out. It's so much easier to say, hey, there's crowdfunding 200 people crowdfunding, 200 projects crowdfunding. I mean, it's like, Probably I will get 50 people from my network donating to this round, even though they have no idea about any of the projects.
And most of the things actually worked well. There are some little bugs here and there and we are fixing them right now. And then Matthew takes a nap right now because it's 2 a.m. for him. And then hopefully those are fixed soon. And then the communication also goes out that everything is ready for primetime. Um... From everything that relates to Certified, I think kind of like in the intersection between my Earth and certified, some things are not Perfectly working, but like it's because like the one thing that I noticed is when somebody has an account and then changes their email on certified.app, then the earth backend doesn't really know about it.
So there's a mismatch between the emails, but that's kind of an edge case that probably not a lot of people run into, but that's something that Ken will just figure out at some point. So, but like in general right now, it all seems pretty nice. I don't know, Adam or Karma, if you have stats on like how many more, if there are accounts created, if we saw already a spike, I don't know.
I didn't check out the spike, but I checked out total accounts now on certified.one with valid actor profile.
Nice, nice, that's nice. We also have a A secret soft launch at 4am this morning because I just Soft launch, the update of the Certified app because it's funny because it's like nothing is ever done before a deadline and I really wanted to get it out before the round launches because it's not the critical path but if people actually go from MyEarth to check out Certified I want the new landing page I think that was kind of the main thing but then the app is also there with the Magic index that actually like made quite a little bit of trouble yesterday.
So I'm very excited when we switch that to the hyper index because I think by now I have like thought too many things on it and it's not actually like clean anymore at all. So we can talk about that.
So here it's just an announcement. Christopher is also joining. Christopher's joining with 50%. on the 13th. So I also had that conversation with him yesterday. So... Things are happening and I think it really feels good from my perspective because what I think really needs to happen now is we are creating right now some momentum with myEarth. And right now is the time then to lean in. And I think we have also seen over the last months that we didn't have enough capacity for the things that we wanted to do and like we didn't set it up as much as we wanted to.
And now with and then also Christopher, we have set up better for that. And with Christopher, we would, Like his focus was moreBe a little bit on kind of, also helping with kind of the business development and kind of the use cases where I'm talking with people about, why is my camera not? Okay, it's not very sharp, but I guess you know how I look. And so basically also in some way, I don't know, there's some term for that frontline engineer or sales engineer or whatever, that kind of position.
Um, Warning to everybody else. I basically also ask Eve to ask all the questions and criticize and give feedback and everything that she really observes. Because I think we all know that we did things sometimes Faster than we wanted to or we just didn't have the ad protocol knowledge where we kind of like tried things out and did it and some things just like We carry it with us and I think that's like totally normal and having the new eyes and Exactly.
The Boomi search launched on July 3rd. Huh.
That brings me to I haven't shared the emergency contact details with the Marath team yet. I need to do that.
I think I also need to write access to Notion so I can add mine.
Yes, and... I also checked in general, Adam will be at his 50th birthday, so hopefully we don't need to disturb him, but he will have his phone with him.
Karma is also travelling but also in general emergency available.
Um, And Um, So I think It's kind of overkill to say like, oh, like, somebody specifically of us is available and like Saturday at 2:00 AM. So in general, it feels like everybody is kind of reachable. And the first line of contact would be the Telegram group or if it's more more urgent signal a DM and if it's more urgent just calling people and that's kind of like the The escalation thing, I don't like if you have other opinions on like how we should do that, I'm happy to hear them. But I feel like that's the best we can do right now also given the small team.
Yeah, I guess. InStatus has pretty nice on-call features that we haven't used yet or even possibly looked into properly.
Any opinions? I mean, it's just simply that you can set up schedules and say, okay, so-and-so is on call for a particular time or whatever and have escalation policies and notification Yeah.
I think we don't need that right now in a small team. I think that the risks of anything going on is pretty low. It's more just like like giving people a heads up that like yes we obviously want you guys to have a weekend we will probably hear of an outage from martz and first because probably one of the donors will start messaging Mars or something, you know, like, like we will hear from it. about it if anything goes wrong and I don't think we necessarily need like a feature or anything.
Yeah, so that makes sense to me. I mean, I would suggest that like if the idea is for all of us to be keeping an eye on things then let's make sure that we're all signed up to the relevant monitors with alerts. Some of us are already but not everyone. That's what I was just going to say.
Yeah, so it's in pretty good shape, but we have had some bad luck within status not detecting stuff when it should have done. Yeah. I'm not sure. I may need to chase up in status support again because I did chase them the last time that happened and I'm not sure I got a satisfactory resolution but in terms of like you know we're not in like it's it's it's a pretty clean setup that we have right now.
I mean, I think the only false positives we've been getting are for the hyper index from, 'cause they ping it from four different geographies and one of the geographies was being a bit flaky, but in general, I think the monitoring, I mean, It could be a bit more in depth, but I think it's fairly decent already. Okay.
But just so you know. And doing less regular work on Friday and Monday as well. But for everything that's needed, I'm there. Um... Two quick updates. I had also yesterday a good call with the 11th Hour Project. That's... part of the Schmidt family foundations, so the money from Eric Schmidt. and I met Sam Oslund In San Francisco, a couple of... months ago and I got connected to him through Wendy from the Internet Archive and he and also Wendy they will be at the web camp and so we will have more conversations there how we can potentially like connect collaborate and potentially also that there might be some funding available as well but he and me are pretty aligned on like let's figure out first kind of what we can do kind of first relationship building and then We'll see what comes out of that.
And then also with the Schwab Foundation, last week was the Board meeting of Bertelsmann where Schwab Foundation put in like a proposal and I have not heard back yet from it but that's also ongoing to quick updates on that and So these were the updates and I put boomy sets in there as well. Maybe you actually want toUh, say a few more words on like what is actually boom research launching and um i know that you wrote if we can also go through it and and check a little bit and play around with it um that's still on my to list um but um Maybe a few words on BumiSets.
And because everything is interoperable, we do want our platform. Sorry, I'm in my admin panel doing a bunch of stuff right now. we do want people to be able to log in with the same account and upload data. And so the idea is that once all of these organizations and projects get funded or are funding, how can they showcase the proof of what they're working on. And so the main big feature is to add observations.
And so you're able to add observations and kind of like audio data and also just tree data and all sorts of other different types of data. and have that interoperable with my Earth, And so this will be the first launch of people being able to log in. being able to connect with whatever same account is with my earth. I start adding in data. We also have a little feed, which you probably-which actually you might have just seen. This is a feat of all.
Like everything that's been happening, a couple of people are testing and Just like taking pictures of trees and that's karma there. We have our community members in the Amazon Parking Assist Tree Boats. They just took, you know, 170 photos. These are real data and this is a real feed from the Amazon. And so this is all real stuff. So the idea is like all of that will be connected to their So you have like a live activity feed of how the projects in My Earth are doing basically. That's the proposition.
If anyone wants to test it, try log in and add observations. Go outside, touch grass, take a picture of a tree in the backyard and let me know how that goes. That's all I ask.
OK, because if it-so--Okay, Karma's saying no. I'm seeing from your head nods. Yeah, because I think it's if you just deactivate, so if part of the association process or if it is to deactivate it also emits an event that the indexer can also pick up and then can just basically index that has been deactivated and that prevents it from being accessed. But you can basically control that through the XRPC method you expose to how the launch works there.
Yeah, so currently the... essentially they're pretty decoupled, the group account on like all the CGS side from the PDS side. and we can implement UI layer Um, papering over the cracks. So, you know, one UI flow that deletes everything and this kind of thing. It can't be atomic. But at the Architectural level, we do have the option Um, of papering over some of the cracks in the way that the the PDS limits admin level operations on an account.
So essentially like the privileged stuff generally, Either you need a recovery key for doing the highest level of privilege stuff on a PDFs account. or you need the email flow of like sending a token to a deliverable email and then using that to grant the thing. Or you can do stuff using the admin only XRPCs for the PDS. So that opens up some options where we can either have a trusted relationship between our CGS and production CGS and although that's limited to just those instances, but that would allow us to do things like allow the owner of the CGS to do admin level operations like sending an email or changing the group handle, all that kind of thing through the CGS.
Or we could actually have the cgs speed accepting email on behalf of the group. And so we set, we have a deliverable email address that goes to a CGS service. like we do with the testing environment with Mailpit, and then we could do stuff that way. So those are just some options and we need some more thought around which of those might be worth doing. Just putting that out there is potentially I think there's other goals like Well, there's transferring ownership of a group so that someone truly owns it rather than it being completely custodied on behalf of the user by the CGS.
There's also things like updating, renaming the handle of the group, like I did for David manually for Game Forest at the weekend.
And that was the thing that got me thinking more about this in depth is like, you know, we can do this as a white glove service, maybe even for the next few months, because it's probably not going to be but it would be more ideal if people could engage rename their, like set the email of the group or rename the group handle, set the group password. Yeah, the point is all of these things are privileged operations that can either be done by the EPDS admin as a white glove service or we can go this other, like these two other possible routes of implementation, which like, between the two or the automated email processing?
So I sort of want to caution against this, like trying to solve a thing.
I'd say give me a few days 'cause I'm still working through the specification and what I could do with it. But there might be something that could help here for doing these things.
Yeah, so on the list right now, we have remove organizations and accounts, deactivations, being able to propagate through indexers or whatever, CGS owner transferring, editing, deleting handle of the group, set emails on the group accounts.
Because I think there's something where if you're needing something right now, then let's figure out how to do that. And then the other things that we-because now we added a feature list for future roadmap of the CGS and potentially EPDES, et cetera. And that should have the right architecture. But if there are more urgent things, then we should just figure those out to make it happen. And like, as long as that doesn't put us on the wrong path.
I see all of these as medium priority that could very well go through the indexers. I would say removing organizations and accounts for me feels high priority, obviously not urgent. But it's like... It feels like such an atomic thing that we want to make sure that like being able to edit a handle being able to have We like indexes having that propagate and so on.
Yeah, but I mean, I think it's just like, we have to prioritize something. And like, I mean, if like the removing organizations and accounts, in theory, this is possible right now, if you're, Um... remove the cgs and then in theory you could just delete the account on blue sky so that's like this also shows me that what we have right now from the infrastructure makes it possible. So we could implement that.
So if that is, for example, like something that you need right now, I could implement that on certified.app so that you can just use that if that's something that's high priority. And then we can look at the other things and make an actual plan for that. That makes sense. Quote. And I feel like those are all things, Adam, that fall into your bucket, right?
Yeah. Hoka, on the certified app, you need to adjust your OAuth scopes a little bit because they're still using the transition one. Thank you.
Yeah, because it doesn't grant access to modify identities. I think, but there is a scope. I need to check with Amelia if it's still there because I know she was cleaning up some scopes that were kind of like not supposed to exist. Okay. And I don't think any of the apps that we built have either.
Maybe I'll write up a skill for it and you guys can use that. So It's probably passed everyone by, but The lexicons have permission sets.
for our lexicon slang. Now I haven't published these yet. on that proto so that I need to do that but so we've got all of the hypersets, so all dot hypersets, app.certified and called hyperboards. those three have permission sets for writing to those lexicons. The intention is that, yeah, like every... every app that we have that needs to write any of those should be using those permission sets and then that's a shortcut to avoid really long permission scopes and also to have user-friendly descriptions of of stuff and the same for API keys on the CGS as well.
I just noticed. But maybe with our topics here we will... get to them anyways. The merging of accounts. All done?
Sometimes when we were trying, like as you know today, sometimes we're trying to merge in accounts. It's actually not a certified issue, but also because MyEarth also has their own backend and database and so on. Sometimes it's just like, we could not use your full karma, unfortunately, because the issue was more like on their end and They're not purely certified. But I think it's good. Bye. Cool.
Lexicons, yeah, but that should be a very minor change.
Cool. Then, Shafi, I don't know, did you have another discussion also with Matthew on the PlusSkill login? Is that, because it's not implemented right now, right? So we are just postponing that.
flow and that's kind of why because I guess you can log in but how does that interact with everything else and also the Ma Earth database and all that stuff So it's just it's not been resurfaced in the conversation because everyone was worried about launch but the last thing we talked about was yeah we need to think this through like having a good we don't want blue sky people to feel like second rate citizens on the my earth app even though they areMe too.
It's like the thing where we need to have that conversation or that they understand the value of what we are doing as a larger ecosystem. But in general, I mean, I Intentia didn't push for it because like having like The launch now was more important than this, but it's still annoying. So this is for me like kind of the highest priority now that the launch is going well.
Yeah. Yeah. I mean, cleaning up data has been basically my number one task in Mars over the past month. First, I cleaned up hypersets. that took a while. And then I, so yeah, I have another call with Ken. It's at three, which is in nine minutes on the latest PR. It's perpetual.
Oh, yeah. I mean, last conversation, we decided that we were just going to market on gain first hand right We're going to label it on Gaten Forest's end.
I think badges is what David was saying.
This is a test badge, and then we filter it from our app.
and also kind of like check out the certified app if it actually makes sense how it is currently using the magic indexer. And if it doesn't make sense, like feel free to like change stuff on certified.app. that I think this would be like one of the tasks for you to to switch from the certified app to the hyper index. And I mean, this is-As long as it works like this, it's not urgent, but like something like for next week.
My break. On the updates on that side, I did complete, I think, almost all of the P1 and P0, and I think it's ready to integrate into certified app right now. The last one yesterday was the endorsement closure. which is, yeah, that's been done. But I think one thing that will... probably have to be done before that is with the lexicons before serving it because currently some of the records that we created before version one of the lexicons are messing with the responses.
And I think your AI pointed that out as well. So right after this, I'll work into merging it into Certify.NET.
Yeah. I'm also warning, I didn't do more changes after I wrote that issue, or like my, I wrote that issue like one and a half weeks ago. If something doesn't work from there, it might also that Certify.app changed since then a little bit. So just check that. Yeah. But also if you don't get this done before you travel, like don't worry about it.
And now we should also what we have like now of the, the open things for the CGS that we discussed, these are all open because they were not on the critical path for Mars. And now we have the time to actually put these things back on like... this is the architecture that makes sense and then in that discussion we should also discuss the use case for the created via or other other things and then I think it could make sense, but let's look at this as like a larger picture and define kind of how we see kind of the this functionality evolve and not just like in an isolation of this.
And I mean, I would put it in draft if that is not... No doubt.
I don't know if I have shed a Rastafari or Kama Probably not Satyan, but I have a document basically at some point I wrote down how to work with me and I think it's actually like interesting and useful. So I actually did ask Evelyn to write one as well if she feels inclined to and in that way I think it would be interesting. This is absolutely not urgent but at some point writing a little bit a similar document and that we can also share that.
And so I will share that with all of you as well. And I mean, I will not make it a requirement, but I strongly encourage you to take a look. It's interesting. And I mean, I also include something like my personality tests because I think they're interesting. But this is not a requirement. There's like in the bottom, there are also resources that explain kind of the intent. And you can basically do it in whatever form you like and share whatever you like to share.
But in that way, like if Adam starts to look at kind of the these things. I think that already makes sense from the CGS part. and And Uh, Karma is looking into the hyper index stuff.
Okay, well let's figure out async what a good day of it if you're not available Monday. Bye.