Hey everyone. Just kicking off from minute or two. Well, it looks like four minutes late. This is Evan. Thanks all for joining. The goal of this is really to talk about what did we learn from that.
hackathon itself. Like what could we have done better? So this may not be the one that we really, you know, pat ourselves on the back as much as the one in which we say, "Geez, it would have been better if we
done X or Y. I can also tease a little bit, and I see disruption Joe on here, and we can also talk a little bit about some of the initial results, but we won't have
too much to speak about quite yet because they haven't been judged yet. So we're still a few days away from actually releasing.
more details on the results in part because we don't want to bias the judging and the judging goes through this Sunday. So yeah, that's kind of my thinking on the use of this time. Feel free to join, leave, comment,
back and I'll kind of go through some of the lessons you aren't as I see them but you know anyone else who wants to come on up and just be us about them or share your insights as it were that would be great as well.
One other detail is if you participated of course and you left us your wallet, you will be getting a Kudos designed by an ODC member today.
So that's good. That's at least a sort of a recognition. And for anyone who's going to be a judge or is a judge, we're designing a special one for you as well. Yeah, I can
share some high level statistics and maybe people will have questions, comments, etc. High level statistics as already shared, if you include all of these submissions, so people that actually hit submit
Whether or not their submission was of any real quality, you're well up into the 80s, which is a big number, certainly a record number for us.
and it matches up well versus some of the fairly recent hackathons from protocol labs and others. So that's great. Caviar, the biggest set of submissions were for the SA. So you're around 20 or so
So there's sort of two essays, but one is really focused on writers, which is great, but I just wanted to make that clear. This is the one sponsored by Trueblocks, which is
It looks to have some really good submissions, by the way, but it's sort of about why decentralized. Where we headed, you know, visions for the future. So quite cool stuff there.
Um, you know, so then if you kind of back those out and keep going, you end up with roughly seven to 10 submissions for each of the
core data science ones that are like during and after civil and grants analysis and then two
specifically focused on passport, which I'm kind of particularly excited about because some of the work we've done has contributed to a better understanding of passport. But haven't we have not had this
focused of a challenge. So you know some good submissions we are at 10 challenges. I actually think it's 11 when you include the data challenge.
That just gives you an idea, right? So out of...
80 mid to high 80s total submissions 10 challenges you end up with you know several on average in each each of the challenges keeping them
to that the essay was the biggest one. So I'll kind of stop. I mean I can just keep talking. You've heard my voice again. Were there questions, comments, suggestions about the...
And we're trying to do kind of a community call on Twitter space here. You know, about the, yeah, about the region, Rangers, Hackathon or anything anyone wants to share.
Not here. Oh, yes. Here we go. Bring someone up.
Hey, Sticker, or do you may be able to speak now or you will be momentarily welcome. Thanks for joining.
right there, thanks for giving me the speak. First of all, I would like to thank all the support during this hackathon.
I need to say it was in a start a little bit tricky to find everything through and also the names like the sandbox which also someone was mentioned yesterday. I mean there's a lot of words that I also need
to learn and looking at pandas and all the stuff and it was a great experience for me. I learned quite a lot. I was participating in the doing and in the after and I made a huge progress.
I would love to say thanks again for the help doing this hackathon and was really a pleasure.
Hey, thank you. It makes our day. So thanks for saying that. Yeah, the
The sort of on board, if you will, like.
I don't know. Some people talk about it as like the first five minute experience. Something like that is on the one hand like really basic and on the other hand kind of hard to get right I think or at least it was for me. So yeah some of the terms.
like even sandbox is something as you said, I think I can't remember their name, but they mentioned it on a Discord yesterday. If you don't know what that is, it's like yet another confusing term, right? And like, is this a data science thing or what? No, this isn't data science. This is just
Approval upon Also just wanted to say thank you to you as well and I'm pretty
psyched about some of the work that you've done also in, let's say pre-ODC, right, around air drop
analysis and so forth. So you know, you obviously brought a lot to the party as well. I don't know if you wanted to say anything about that kind of humble brag about where you came from and how that helped or was or was not applicable when it came to the hack upon.
Yeah, sure. Honestly, it was around months ago when I started going into subalionalizes because I was helping out on an adrop campaign with a
just with a dashboard and then he was a question me, but what if two people maybe have the same amount and bridge at the same time or nearly the same time how it will be distributed
it. So is it by first block time and then I was just choking and saying oh this will only be seablds. And yeah, then I discovered all with there a pattern of a glass ton with 30 wallets.
And then I had a form team about this.
This was, to be honest, my start into civil detection. So, yeah, I need to say I really started into this just a month ago.
So, and then I was coming over the ODC and I found this very interesting also the Heikerton and with Bitcoin.
And so I started with the
a view of this data and when I first looked into, I mean, kid coin is really something different than just say, hey, we make an audio error, here we have
a list of allegoral wallets and can you have a review over this list? I mean, get going for me was a little bit tricky because there even some
funds receiver are even voters. So this means this is then even kind of tricky to really sort out if you have not really a good algorithm or a good
indication or thinking of how you can select them out then they will get false detected. And so this was for me even the first thing where I was thinking about this is complete something else and then also something
is I mean Bitcoin has quite a lot of data quite a lot of users and rotors and this is also not the same amount that I reviewed before but yeah I saw say before I would really love to say
thanks again. I made a huge progress during this hackathon and I'm really looking forward for all the feedbacks. I was wondering do we even get feedbacks for the for our presentations?
Thank you for sharing also your journey.
And I want to come back to data and talk a little bit about that. But yeah, so feedback.
We haven't done much in the way real feedback in the past. Most of it has been pretty informal. That may still work in this in this way. I'm not sure, but I think it's a great point. So what I'm getting at is
what we do and we have this
So judges have an application that they can use as a part of the hackathon application to basically slide scores. I mean, it's sort of like a very fairly simple.
In addition to that, we're trying to capture more details from the judges in the form of basically spreadsheets. And I think that is how it's going to go that judges really have
the freedom to decide whether they just want to use only the DAP or if they also want to provide more decals which is a different speed of preference. If they do the DAP,
I don't know. It's not trivial to find, but you can find actually sort of similar spreadsheets with the criteria from all of the prior hackathons.
And so that so at least level detail and long story short, I think we are going to be able to share which is pretty cool. But in addition to that, we're trying to do more around
I don't know if we would call it mentoring or something like that, where we basically similar to the discourse, discourse session we had yesterday with Storm. And so maybe we should do something like that where we
you know, ask the judges or at least me, you know, can help summarize what the judges said. Some in, you know, telegram chance with them. At least I'll do that. But maybe we can get some more of the judges.
One of them is notoriously media shy as this disruption Joe will know who I'm talking about here. So I'm not sure, but he would be amazing if we can get them to get our work.
Maybe in public with us and kind of do an office hours on I don't know, but what are your thoughts? What would be a good way to get more feedback? I'm kind of rambling at spreadsheets on the one hand second hand, at least someone like at my level is more about
high level overview of what we saw within an addition maybe we have an office hours discourse session on with a handful of the judges. What are your thoughts? What would be useful to you? Most important question.
I think the spreadsheet would be enough, but yeah, I mean you cannot force them to give feedback. I mean, it's still free, but as long as there are some kind of feedback, maybe it's always nice.
Yeah, yeah, exactly. I think get all.
I think we will end up the other thing that I've put into the kind of document, which is very, you know, just a Google dot on the discord is
also making sure some of the results get used. And I mean, some of the nature of open source, as I think we all know, is, you know, sometimes you run into somebody months later and they're like, Oh, yeah, I'm using that every day. And you don't necessarily know
know, right? It's open after all. So it's likely the algorithms are getting used more than we recognize. But in particular, there is one sponsor, i.e. Gekoyne, who
has a couple of related projects where we now have more engagement. I mean, I think we saw Jeremy as an example on our Discord channel. He now leads Passport and we're chatting with the person
was just brought in to lead another product within Kekwain, which is called Grant Stack, which is basically an application written on the protocol to help grant operators. So we're having those sorts of conversations. Anyway, the ultimate
back is that somebody uses your insights. And so we'll be going through the results and kind of spoon feeding them as well to a variety of use potential users like Bitcoin. And that may also spur
the opportunity for additional discussions depending on what project etc. So it just as an example there's a grants back to sandbox. There's a grant stash for sandbox project which you know folks are working on. That may end up being or at least underlying
algorithms, you know, something that gets used or at least suggested for use in future grant reviews just to pick one. Or some, you know, some of the subanalysis you've done, some of those approaches may be those are things that get built into a possibly. I'm just
I don't know. This is what I would do if I were on their side, but I'm very biased towards data. But it's the kind of thing where if I were writing a runbook and I wanted the next 100 users or 1000 users of the kick-home protocol and passport itself,
To succeed I might you know hat tip point out point them towards here's some example folks region rangers i.e. Who do you see numbers? Here's some examples of the analysis they did so that's the kind of work that all
be doing and others in the community will be doing to get adoption up and that might result in further conversations right if people are interested it's got to be a two-way street.
Right? Because as, let's say, there's a future user, I don't know, I don't want to put anyone on the spot in this call, but there's definitely some folks on this call who know this space really well.
or supporting other projects as they come up with what their grant program will look like. It's possible that we get more engagement directly from projects or algorithm creators or sandbox project creators and yeah
the actual grant programs or the firms, experts, consultants that are actually doing projects for to help build grant programs, whether they're, they get get-quain employees or non-get-quain employees.
So that was a bit of a mouthful, but did it make sense? Yeah, I don't have to keep, you don't have to stay up here, keep talking at you, but I'm really talking to the whole crew and feel free again, anyone who has feedback, suggestions,
about the open data community, region rangers hackathon, raise your hand, we'll get you the mic. You know, pods here just for a second to see if that mouthful made sense.
Well, again, another thing I would point folks to is if you have some ideas, you don't want to put them in, you know, the Twitter, put them behind the Twitter event horizon as it were, because this is being recorded, but you have a minute or two.
or longer, there is this dock that is set up. It's just a Google dock, so Chamon me, Google, centralized, etc. But it's in, you should see it right in general in the ODC discord. And feel free to put
thoughts in there as well. And again, it may be sort of a marketing faux pas because we're focused right now on what we could have done better before we've even bragged about some of the very cool results. But maybe I'll talk a little bit about those. One thing,
I would say is a common theme in the open data community. Hackathons, of course, is the freaking data. Right? So not surprisingly, if your community set up in part to democratize data analysis,
Well, and data analysis for good is that we're the problem statement in part is that it's overly hard to do this kind of data analysis. And part of that can be getting access to the right data.
So I open to feedback on on that.
What I found this time around was, look, we took a big step forward. In the past, we kind of were like, hey, let's people use ocean and park because it's a way to use a decentralized underneath IPF.
or, you know, our weeds, or some more storage layer, and then do some community level individual controlled access controls on top using primitives like NFTs basically. But what was the killer you
use case to be open in the last hackathon. Not sure we really nailed it. I mean, yeah, I'll put my results up there, but how exciting is that to people? I don't know. This time we said, look,
We have some civil lists now to the point made earlier.
predicted symbols. Right. So let's use stats 101 to understand that these are not, we're not, you know, trying to put a black mark on wallet forever. Nonetheless, there are lists of
Folks that were wallets that we've identified as Symbols, many of which were inherited by FDD, from which is again a GITCoin acronym, but the data science team within GITCoin
and so on. And so we put those up on Ocean to give everyone a set of really ground truth, if you will, to tune models if they wanted to use them for that.
some of the contests. And we also made a commitment to continue to do that. So as more syllables are found, we're continuing to put them up there. So that's a big step forward. However,
It introduced a piece of friction in an additional piece of friction in the process, which is to be whitelisted is somewhat of a manual process, which is to say fit.
The platform itself is not integrated with ocean. And so what has to happen is an individual has to give us the role it and you don't have to.
when you register for the hackathon. So again, we didn't design that piece of the workflow. And so we had to email people multiple times, ping them to remind them to give us the wallet, it's a source of friction.
Once we have the wallets, we just need it on a script, not that big of a deal, to add them to the white list for ocean. So the combination of those two things probably slowed us down and it definitely introduced friction.
And if you have data analyst who has so much time and they want to get to AHA as quickly as possible, just like we all do when we're trying something out. And like, shit, I'm not even on the white list. I can't even see these civil lists yet.
Hard to know, but probably had a negative impact on yeah, you know probably cause some dropouts etc. So that's something that
We're keenly aware of. I can't say that I have an answer to it right now. You know, I think again, an all things better communication could help.
better explanations. Okay, what is this hackathon? Oh, okay, more bold letters saying please.
enter your wallet, et cetera, that may be helpful. Yeah. But there may be some way we can zap here, the heck out of this thing in the future as well.
You know, the build box platform itself is evolving. I mean, initially we didn't get any. The data layer of that is evolving, which is put it that way. And, and they're very responsive. So it's
It could be that we can just add something that enables us to API-like basically, which would be great. So we could zap here it or otherwise loosely integrate it.
Yeah, that brings up the overall, you know, ETL kind of automation question, which is a big one and curating data sets. And I think this is an area where we probably need to put something in process for.
between the hackathons, right? Certainly the seabulls, this will be the case.
but looking at what can we do over the next 90 days before the next hackathon or less than 90 days so that again the zero to one initial moment is faster.
And there's going to be it's going to be a better data plan. We did have we do have a written plan about how to use for example, I saw it the read earlier from true blocks, but for example, we
I have a NoDC member who has a, I believe an archive node or a couple of nodes running. We want to make that my hands on laboratory, including crewblocks.
and really point people towards.
We're going to point them there, but you know, at least let them know that we have some resources there. That was intended to be done before the hackathon, to be open, but real life, you know, got in the way. And so that is coming along.
But I just welcome anyone who wants to participate in the data layer. I know there's Sean and others in the community who are working in that space. Apeochi and others. So that's something we can improve.
I think we need to market them specifically. We put it that way. And this is a public good where, you know, if you're an air, you do an air drop in someone like that.
You start the stacker you do. It says, hey, you're 78 or 780, whatever number is, Sybles. You know, maybe we want to take on the responsibility of ODC of kind of curating them, adding some metadata, IE,
where they came from, what were the techniques used to gather them, and then putting them up free forever on IPFS or maybe they're already there, but then also access controlling them on ocean. And that's certainly core to what we're
we're going to be doing over the next year. But I think, again, communication. Maybe there's a way to, I don't know. We do a Twitter space or a discord or something on. Do you know where circles go to?
you know, to die or something. That's not the right way to put it. You know, really try to shine a light on it. Very open ideas there.
anyone who wants to come up, you're welcome. I think people are multitasking across multiple meetings, which is or otherwise, you know.
Not a whole to come up, which is fine, but just want to keep reiterating the invitation for those who are interested.
Just one more point on the data.
I think one area where it kind of harmed us is we did not, I mean on the one hand, the dick coin side, there's more data, right? There's a lot of data. You have the cost of all these data sets, but then you can also get a lot of the data on, change yourself and then you can also
So, okay, you've got password data, but I noticed one of the, at least one, maybe multiple respondents, participants, hit the GIT coin API, tens of
of times to get kind of up to date info on passport, the passport API that is. Anyway, so you've got a wealth of data there. We could probably do better at communicating about it. That's great. But then on the other hand, we try to spin up this
trading competition because we have folks like friends at Pantera or elsewhere who are like oh yeah you know when we're looking at loss trading what do you think we're doing in part and also this may be more applicable also on the NFT side
Like we're looking at whether all of these transactions are tied together, i.e., whether these wallets that initiate these transactions are tied together.
to the gas network analysis and other pieces of analysis that have been done and that have created these Legos in the past.
But we provided zero data. We didn't provide any curated, you know, market list data. There actually was at least one or two
data sets contributed as part of the hackathon that may provide a good foundation for future analysis, but we did not do it proactively. And again, we just ran out of time.
time or didn't prioritize it. And I think that's why we got basically no good entrants in that space, unfortunately. So I think we've asked around people
your everyone's feedback is welcome here. The initial feedback I've gotten is like, oh yeah, the premise makes sense. The premise being, there's some over like behavior clearly here and coordinating
Some of these we are illustrating some of this boss trading and so that seemed like an area of growth for us.
And obviously, DeFi is still even bigger than public goods funding. Amazing. To think about, but anyway, so seems like a good area for us, but our execution, we need to even, we need to better
Yeah, yeah, curate data, but you know, probably overall that could be a priority for us, you know, between now and the next hackathon is really
make a concerted coherent consistent effort to get to if we're going to do a similar kind of
challenge next time, we want to really lay the groundwork. And that includes data sets, but it probably also includes attracting more people from that side of the world, if you will. Yeah, and it includes all of the educational
materials. So just a comment on that. It's related to data. I think the sort of Achilles heel there was not having curated data of likely, I don't know. You know, we could literally, for example, I know there's some data from Bloor that was donated
I have not even looked at it, but I just noticed that in one of the submissions. So, you know, having something like that ready for a user in the future probably would be useful as my thinking.
This is like one of the weirder Twitter spaces because I just seem like, you know.
sort of self-criticizing into the void. I think it's transparency all the way down. But it's not to say that I'm not proud of the community and the work we've done. This is yet another hackathon that is bigger than my last.
We tried a lot of things, some of them got real traction.
again I think we're close
I think the pipeline from the work we're doing to usage is sort of shorter than ever, which is great because that will then give
support as a community going forward, all those things. And yeah, I am also hopeful that the governance work, which is basically hats protocol and getting
And I'm not going to be talking about how to do that correctly with folks like Hedgy and others.
All of that, putting a place of more solid legal foundation, all of those things are going to help us spend the flight wheel and the success of
around us. I think will help us as well. So I feel quite good about where we're at. That said, you know, you're never
more clear on what could have been done better than right now probably. So we got to capture it and learn from it. It's my take. Yep. I think
I've covered a lot of what I wanted to cover. I'm going to look at this written doc right now just to see if people have added anything. I'll look at my own comments there as well.
Yeah. I mean, we could do some analysis. I think this is something I've experienced in others do, but we could do some
analysis of the actual hackathon adoption awareness pipeline and basically
how many folks register an interest? How many folks come to these sorts of events in the run-up, etc. So that kind of analysis will be useful. We have historic data, so it might be useful in the future, so that's one
of the comments on the document and why that's important is probably pretty obvious but it just helps focus efforts. Keep in mind that right now we're all volunteers. We hope to have a little bit of funding for at least bounties and maybe
Bouties between the hack-upons to continue to build capabilities that looks very likely. Yeah, the breadth of challenges, that's a question, is there's pluses and minuses
there. I mean, 10 or 11 challenges. It's good, but maybe we want to have, I don't know, three deeply technical ones, right? Where we
I try to get more of a critical mass. I don't know in a couple.
And then a few that are less technical. That's another question. We obviously got a lot of participation in the SA.
even though incidentally TrueBlock set it up as like there's only three winners which is which is interesting all of the other bounties are set up as
more like a pool that we can decide how to split. So there could be more than three. But anyway, there seems to be real demand for random
interest in these sorts of non-technical or yeah in that case totally non-technical. So do we follow the data and double down there?
I would assume we would want to, but that's something we can kind of design in public for the next act of on here.
Yeah, it's our just scanning through.
These discord sessions we talked about it maybe for feedback
The last one, which was just yesterday, seemed to go well, but then in addition during the Hackathon, Palpal had a couple of these.
That's something we could maybe program as well.
And then maybe there's...
something we should do with, you know, the dashboard building communities.
I'm not sure. We definitely did some Twitter spaces with or one at least with June in the fall. Again, we didn't get as much participation
but that was the first hackathon. So maybe that is something that we should consider or metrics down or similar.
Because that's when most people think of data science and what three, they're kind of thinking at that level, which is fair enough. And frankly, the rounds need dashboards, even those sorts of, I know Umar from Gekkoing kind of holds something together on that.
on snowflake or flip side. And that's great, but we probably need more dashboards. Maybe, I don't know. Dashboards. Anything else folks want to chat about?
Yeah. I've just gone through the list such as it is. Maybe I'm too close to it. Maybe we are too close to it to really know. I can go through some of the just keys, some of the interesting submissions of people.
You know, people are interested in that. I don't know. I see somebody from one of the sandbox projects here, so they may have dug in deeper, but we did get a number of contributions to sandboxers that phrase again.
which are basically just software projects that we are supporting, giving a little bit of attention to and other support, such as including the Minhaqa Fund. We did get a number of submissions. One has to do with LEGO Curation, which is
You know, sort of a meta challenge, if you will. It's a yeah, which are these legos is similar, which are different. So it's that zero to a high moment where so ODC you say you you curate these algorithms as well.
you help them get pot of those legos. Which legos should I start with? Presumably they're relatively educated users so they look at them and be able to grow up based on
based on the underlying data science, the underlying algorithms used. This is basically cosine looking at the variance.
Or this is a different approach. But nonetheless, LEGO curation seems very important. And there is at least a couple of contributions there, which I thought was a good sign. So stay tuned for that. This is a project that
It doesn't really have a single owner right now within Node.ec. It's kind of strategically important. So we kind of created a space for it.
it, IE or repository, etc. And it looked like it got a couple of contributions, which could be great. Yeah, just looking around at others.
Yeah, the on-chain passport stamps. I mean, there's
There's a couple of submissions that look couple look like kind of no brainers, which is good. So.
Yeah, it's possible that that will actually turn into a stamp here. I think one question that we have or I have is
In all of these things, there's a spectrum from do the data science to make it so. What I mean is, maybe it's obvious, but exploratory data analysis on the one hand, figuring out, okay, based on these clusters,
etc. How do I then formulate that into an actual passport stamp and all the way up to I did the pull request it got accepted there is no a stamp you can see the stamp it's running it's so that's make it so and
You know, frankly, should those be broken into multiple, that's something that we
are doing a bit of into multiple steps that are rewarded. Like, use an existing logo.
or maybe as an example, take existing Legos decide what should be a stamp or that kind of thing. Or should we just put more money in or otherwise have fewer challenges.
and make that all one big challenge. You probably won't win unless you do the more boring non-data science stuff for data scientists of getting it upstream. I don't know, it's a
It's an open question. I think given right now that most of the submissions are single individuals, not teens, you're now asking that persona to really
I do more than one thing right now. I was just thinking about the data gathering as well. Again, if we don't, depending how much you
spoon feed the data you could need on the ideal team. Somebody who is a data ops person, data engineer, let's say, data scientist, and then software dev.
Frankly, you can kind of do all three. I mean, it's I don't know that it's not like they're you need a PhD for them. But nonetheless, you know, we could have personas. The ideal team could require three personas if you're going to have a challenge that goes from soup to nuts as it were.
Or again, do we split them into three and this is you know, you're on the dev track and maybe
giving future hackathon we have.
So that even when you use build bots, it's basically a two-level hierarchy. You can have tracks. And within tracks, you can have challenges.
So thinking out loud, we could have software development track. We could have data engineering track. And we could have data science, deep data science track. And then within those, yeah, you could have a couple things that might be a way to
to kind of pay by numbers for the personas coming in. They can see clearly more self-identified over here. No, it's an interesting design space and nobody's ever
anything else. I'll just I'm just taking shorthand notes here. Yeah, I think should we
make tracks based on personas and then challenges are specific. Based more on the needs of
sponsors or pain points to be. Well, I think I'm going to let everyone go. Thank you all for joining today.
I hope this was at least some interest. I appreciate you spending a minute or two or an hour in this case. These things do live on. So this did become a little bit of a learning in public. I think transparency is important.
First we all think that, hence the open data community. But yeah, I don't know, StackVarue, dude, did you have any closing thoughts? I keep putting you on the hook, but thank you again for joining. Thank you. I see you, Bapescu.
for a tune. Thanks again for joining. Yeah, and any last thoughts? Yeah, go ahead.
Yeah, it was great to listen to you. So, it makes also now for me quite a lot some stuff more sense and I think this needs to be really in the onboarding to really onboard people directly. But yeah, I know onboarding is quite a difficult topic. I was
working earlier also in the dough. So I know it's sometimes difficult to bring over what you mean that does us really understand everything. But was a pleasure to listen to you this hour. And the last thing I was thinking about maybe even a feedback form for participations.
for possibility or so they can give you feedback back. Yeah, we did a kind of a survey last time. We need to spin up the survey or something better. Yeah. And maybe we can do it over the next
If you'd have it ready by when we announce the results next week, so that'll keep people motivated to actually give us responses if they find out whether they want. But yeah, that's a great point. I think we'll do that. All right. Thanks again. Thanks, everyone.
Have a great rest of your day.