0:00
yeah yolo we're actually looking
0:11
[Music]
0:20
hello and welcome to game of nodes a weekly podcast on the cosmos from independent validator teams
0:25
and as you can possibly tell by the look on our faces we're pretty
0:31
tired this is not coffee it is for null i guess because it's like
0:37
because it's 708 a.m it's morning in kanga time isn't it whereas it's night
0:42
in europe time and it's apparently one o'clock in uh
0:47
american time yep it's after five somewhere and that that somewhere is right here it's always have you ever
0:53
heard the uh i've i've been on quite a few stag days this year because people have been getting married off the pandemic
0:59
um and they have the the thing that jumps to mind immediately when i see booze that i don't know what hour of the
1:04
day is for you right now but is uh one before 10 or 10 before one
1:14
i'm not sure you should be drinking before 10.
1:19
[Laughter]
1:24
i love the way i love the way there's even there's even a specific joke there are two specific joke sound effects but
1:30
it's it's got to be the uh um i don't know was that me
1:37
so funny you don't even know it i don't even know um so yeah i mean uh
1:43
there's there's mercifully i was gonna say there's merciful little drama to talk about this week um
1:50
but i guess there is there was one thing i suppose has happened last week that maybe i don't know if you you want to
1:56
talk about any of the juno growth fund stuff um it's very juno specific though but i
2:01
guess we do kind of cover quite a bit of junior stuff and it was quite a loud loud noise although it
2:08
i don't know i've been so busy this week that literally i was working and i saw that there was like some noise and then
2:13
by the time i stopped working and was like what's this drama the drama was already over
2:19
so it must have been a bit of a storm in a teacup is my feeling on that one
2:25
can you break it i don't even know what you guys were talking about what was what was that uh so there these both happened in this in the last
2:31
week right there was um junior growth fund which is uh the description was
2:38
basically the junior growth fund and lube where the two things have happened the loop one is the one that literally
2:43
was was over before i noticed thanks to working on aptos for like the last 72
2:48
hours without sleeping um but the junior growth fund is more interesting that was the one to
2:53
basically would draw some of the prop 14 rewards from juno swap pools in order
3:00
to fund a vc like fund for juno to incentivize new projects basically it's
3:05
the continuation of the terror developer fund okay but not for just terror projects okay um
3:12
and yeah what's interesting about that is like i mean so i think all of us have kind of sort
3:19
of bootstrapped ourselves into this space like we kind of ended up here a bit accidentally and we've kind of funded ourselves until we could make
3:25
some money and stuff so i guess like i'm i have elements of skepticism about the
3:31
vc model being a healthy one but also i guess recognize that there are
3:36
some projects that need more investment up front that need larger teams to get stuff done and all that kind of
3:42
thing and so it kind of seems like an obvious good move even if it's not really like the way that i roll you know um
3:49
do you think that teams lose the hustle by doing that do you reckon like you know they just they don't have the grind they don't
3:55
have the they haven't done the tough yards they don't you know what i mean like they haven't lived in a caravan and uh
4:03
haven't lived in the caravan who's shot in a pot yeah and eaten baked beans for like seven months while they're trying to get
4:09
their thing off the ground do you know what i'm talking about yeah i mean but paid devs all straight up as well
4:15
like you know i'm hiring a team we're going to build a thing everyone yeah i mean
4:21
i mean i guess this is like a cultural difference as well between a lot of the terror projects and like the stuff that organic more
4:27
organically happened on juno is like a lot of stuff that more organically happened on juno and maybe in the wider cosmos like people
4:34
had to kind of generally speaking get their thing running and work out how it could make money
4:40
as part of getting it off the ground whereas a lot i think maybe if you don't do your funding if you get funded ahead
4:46
of time you may be disincentivized to work out what your funding model is um
4:51
and that's how we end up with advertising right in web 2 which is like incredibly toxic and corrosive as a
4:58
as a funding source so um but i think the vc piece is healthy because
5:04
there's the part that is healthy about that is is when the numbers are realistic vcs also look at who the team is what
5:11
that track record is they have enough experience to see if the idea makes sense or just a bunch of [ __ ] right
5:18
um the community does not have that experience so anything that sounds good like oh yeah
5:23
and you don't you know let's say you don't know who the team is behind that you don't know if it's a real it's a real structure that they have the devs
5:28
to do it where it's just you know i'm just throwing cash at an idea and the idea might be good it doesn't mean that that idea is going to come to light
5:35
right the outside thing it's like there's no due diligence it's just
5:40
[ __ ] yeah that sounds good i i mean like ever that's the thing like everybody that's why everybody votes yes
5:46
and most proposals is because we want things to go right we got the rocket emoji going like [ __ ] crazy so as
5:52
long as yeah right they want you they want to see development they want to see stuff right on the chain but they
5:58
don't care about looking at what it is or just everything right and we talked about
6:03
this i think we we talked about this back in uh i don't remember what it was but we talked about this i think relation to
6:09
was it cosmo verse and somebody like all the props coming together like hey i wanna i want twenty five thousand
6:14
dollars but banners up at cosmo versus like what the [ __ ] is that like where where's like a real marketing budget like show me how you're going to use
6:19
that right i think it's 75 000 per chain right per chain right making the rounds
6:25
right yeah yeah and then one after the other is not even all at the same time like that's it seemed a little bit um
6:30
you know shady uh well not shady it just it didn't seem like the right way to do it
6:36
like i feel like the right way to do it would have been to say okay we're gonna ask for funding from seven different chains or
6:42
whatever the number is all at the same time rather than like one after the other oh there's another co-sponsor oh
6:48
another co-sponsor like i feel like it was a little bit disingenuous as to what you were getting
6:54
like it's like oh is there two co-sponsors or is there 14 like right
6:59
yeah no like no um so not harping on about that and going back to it it's a
7:05
good event um value for money is uh
7:10
i don't know i don't know the outcomes i haven't had like any um
7:16
you know feedback or or analysis of like
7:22
what it has what the outcomes have been for the for the chains but it seems expensive for the amount of people that
7:28
attend it considering they also buy tickets and it just seems like it wasn't going
7:34
about the right way but hopefully um there are some lessons learned and these guys are new as well at um you know
7:41
they've only done one or two events so um they're new to the event game um but hopefully uh in the you know next year
7:48
they've learned some lessons from this year um and maybe do it better i hope it'd be nice right
7:54
yeah so but so back to the that's the vc question though i think please point out something like
8:01
i guess interesting which is the the question like who's doing the due diligence which i think
8:07
uh like obviously i kind of yeah i think that the junior growth fund's a good idea like kind of behind that
8:14
um but like the question of like who who yeah who is qualified to do due
8:20
diligence i guess it's that thing of like there's a community pool there which is just anybody can put up a prop and ask for
8:25
money from the community pool sure but the danger of that basically is that the community well it's not so much the
8:31
community the military is malicious towards the
8:36
community fund kind of the opposite you get very few props related to the community fund which is interesting i
8:41
think um but also the yeah the community maybe is not the best place to do
8:48
due diligence or communicate around due diligence but so this is effectively sort of
8:53
taking a portion of that responsibility out into effectively a multi-sig right yeah but i
8:59
guess the question is like you know how a lot of this like kind of comes down to okay so who is qualified to do due
9:07
diligence number one two those people probably don't
9:12
aren't in the junior organization already in so far as there is any organization even even if it is very scratchy and
9:18
it's multiple telegram groups and all sorts of stuff but like a lot of this to me comes down to like
9:25
juno in reality like the core team is quite small right and it is like it's just like
9:32
every other startup i've worked for all be a different level of chaos because of all the anons and different telegram
9:37
groups and different sure people working independently on stuff
9:43
but the problem here is still one of scaling a small team like that's the fundamental problem is we
9:48
to do vc type stuff we need um you know due diligence and those sorts of skills and we need to scale a
9:55
team or juno needs to scale a team yeah and that's like really hard that's like
10:01
literally why all organizations die or like small teams or startups whatever they fail because they
10:07
find it hard to actually like work together when they have to go beyond one or two people that kicked off the whole
10:12
thing who implicitly trust each other because they ship the first thing um so yeah it's that kind of growing
10:19
pain that i'm not so sure about i guess like the rest of it i don't think i think it's a bit of a no-brainer
10:25
yeah and i think i mean i'm not not trying to i just asked about what the heck the dao was so i'm not
10:30
obviously have no opinion on it but the idea um i like the idea and it'd be
10:36
good i think there's the one thing that i would say in some of these types of things is i don't know i don't i see a lot of
10:42
accountability meaning like after those things go forward and after that decision is made does anybody actually
10:47
stop and look back and say hey was this was this money well spent or other types of things and like use that knowledge going forward it very seems
10:53
it's very much point in time where an organization that's kind of built around that would have a portfolio and they'd
10:59
be measured on the performance of that portfolio right yeah meaning like do we make the good investments and are that we have the
11:05
right people in these roles to make those good investments and so if that dow does do that meaning like we we have
11:10
some ability to do some due diligence and we have the ability to look backwards and say this was smart or this
11:15
was bad or why do we choose this and like actually build a repeatable process that can get better over time
11:22
than hell yeah right i mean i think i think it would be nice if if those if the dials would post like yeah like a
11:29
long form report of what's what's happened or what's not not what's worked what hasn't worked right um yeah like a quarterly like a
11:36
quarterly review yeah and maybe there's a service that's going live very soon that allows daoz and
11:43
multi-six to natively post as if they were on a social network maybe there is
11:48
a thing maybe there's a thing yeah i mean maybe someone will finally
11:53
finish that thing that they've been talking about for months and months maybe it's maybe it's already
11:59
maybe it's already finished but the hard part is getting legal side off and all that crap um
12:05
i feel like this project like hasn't even really been doing anything
12:11
yeah no it's not all the work but but like yeah i know scaling stuff
12:17
is hard i guess and actually like um the the question is like was you know
12:22
this actually so i think i've see i've essentially called this like the burn out cast or something like that because we're all
12:28
pretty shattered but i think like so i was thinking about this in the context of um
12:33
so that there's a couple of things like there's the juno growth fund which is uh you're going to get funded
12:39
somehow which the relation to that was you know reducing the incentives that were
12:44
applied to some of the genus swap pools a potential way of getting some initial funding into it um
12:50
and then so you know there's there's a scaling problem which is scaling juno's ability to
12:55
do the funding yeah uh interestingly it touches on juno swap who are having a scaling problem i think
13:02
of their own at the moment in terms of team and then i think we're all really really
13:08
tired over here in validatorland and i think a lot of our validators i think a lot of us are basically doing
13:14
the work of several people um and so there's also it kind of seems like burn out and scaling being hard
13:22
like these two problems like i'm like just the last two weeks that's all i've seen every team every
13:28
problem has eventually devolved down to there aren't enough people or we don't have the right people we
13:33
don't know how to get them right now or whatever like it feels like that is a really common problem and that's like
13:39
underlying both the tension in that prop well that commonwealth discussion
13:44
and the funding itself has a question of like how do we scale that organizational tool
13:50
and the reason all of us have bags under our eyes is because we we run validator operations that we've not scaled yet
13:56
right so yeah why what so what's with that
14:02
so what you're saying like the fundamental problem is is that there's just not enough people um in the
14:07
ecosystem particularly around developers like everybody's looking for for go
14:12
developers to work on the sdk in their own projects and there is just no one
14:18
and i mean even like juno's looking for people we know a
14:23
bunch of test nets looking for people you see
14:29
someone mentioned calum um so
14:34
yeah i mean you guys are probably um once you launch you're probably going to be needing um people there's
14:41
god i'm looking through discord just here and there it's just i know there is a number of these projects that are
14:47
trying to find extra people um and you know what that means is that
14:53
people who are already working on projects are you know out of necessity to help those
14:58
other projects like going across and lending your hand there as well and then they're becoming more
15:04
stretched and then you know they just get tired and overworked and then non-productive and
15:10
[ __ ] to bed and have to have two weeks off and disappear and don't talk to anyone on discord or anything it's like they
15:17
don't actually have friends or anything like are you recognizing anybody in this in this
15:24
you know make everyone else feel lonely because they just disappear and don't talk to anybody
15:30
for the record i took i took literally five days off and then i had to work for two weeks like 12 hours a day to have
15:37
everything squared away that i could take five days off i still had to work a morning and those five days off
15:43
and then i've been working 12 hour days since i got back i mean i haven't seen [ __ ] posting from calum either i mean
15:49
that's been disappointing yeah i think callum's been taking he's he got very very burnt out not from the
15:55
rust but from the [ __ ] posting like it's quite it took i think quite a toll on him you know keeping up that cadence of [ __ ]
16:01
posters mentally taxing doing that one liner a
16:07
day and then a couple of like witty comments i mean i can see how it'll work yeah
16:14
there's a lot of work you don't see now you know there's a lot of work you don't see like it's the it's the torture like
16:19
the anguish that goes into that that moment of inspiration do you ever watch that old british um
16:26
like animated thing called monkey dust like really really dark black comedy
16:34
it's like a really dark like thing that only ran for a couple of seasons i think the creator died like really young and
16:39
that's why it didn't you know whatever you can get like a dvd of the first series i think that's literally it um you can find some bits
16:46
on youtube um but there's like this there's this thing about you know those like really [ __ ] catchphrases you see like big
16:53
businesses have where they're like get inspired or something there's this whole sketch about how
16:59
it's just like we need a new slogan and there's like a bunch of people in the boardrooms like god damn it renshaw we need a new slogan
17:06
i don't care how it's done and then they they go like well i'll get our top man on it and basically has this guy like
17:11
goes out into the wilderness and then like disappears for several years and there's like there's like him getting
17:16
like mugged in the desert and then like beaten by beaten by a gang of men and stuff like this and then like he's like
17:23
years later he's got his long beard he's like god why have you forsaken me and then like it shows him coming back
17:29
into the boardroom and he's just like yeah the new slogan is ignite for the future or something like that like the
17:35
punchline's always [ __ ] but like he's suffered so much for the crap and
17:40
there's like a series of these there's like one more they need a new car name and he goes and like yeah he goes on like a spirit journey
17:47
and then ends up in like a german sex dungeon getting like just having like gang sex and then while
17:53
he's having this is a very dark a very dark and quite explicit cartoon i should say it was on very late at night and
17:58
then there's like he's basically um he's basically getting beat like i think like you know for sexual gratification like
18:03
getting beaten by a [ __ ] or something and he just goes ah lisantrano and then and then the next thing you see
18:10
is they're back in the boardroom going excellent and they're like unveiling nissan toronto
18:16
something like that uh they have one where they're trying to work out a new name for cancer like is all these this but the the long line is
18:23
like it's always a [ __ ] result but he always suffers so much for it and i feel like that sketch is kind of like how i'm
18:29
[ __ ] posting i just except maybe without all the sexual explicitness but i mean he did dial into
18:36
this from a bed so you know what are you gonna always give me bad knobs of broomsticks though that's a you know
18:44
relevant thing god i'm trying sorry sorry sorry to the listeners i really need to emphasize
18:50
that i was up until like 3 30 this morning working on um working on bringing up some new
18:56
validator stuff so it's been a it's been a very long few days and terrifying right
19:01
wow yeah well so yeah so i've been going on a bit of a
19:06
spree anyway on all of our cosmos stuff with automating everything with ansible um
19:13
so yeah because well you say polka chu friend of the podcast
19:19
legend um has a bunch of terraform terraform ansible stuff for their operations and
19:24
so i've been basically adapting some of that changing some of it write some new bits for the way we set up our servers
19:31
it's gonna have like a bunch of new services come up on pokachu.com
19:38
yeah straight straight up forked um but yeah the um yeah so some we're basically so we're
19:45
looking at apptase which was the outside of cosmos so i don't want to talk about too much but um they provide a bunch of um terraform
19:51
and stuff as a template to work with and since we're aws anyway generally um that's actually quite a good fit for
19:58
for us doing something that's very repeatable for a new engineer who's not me who doesn't want to like log on to a box and go this
20:06
has all been configured from the mind of one very tired person like if the foundation has a
20:12
recommendation for how the services should be set up i'm happy to use that as a template and
20:17
you know just add the additional layers of automation and stuff on top that we'd use so um
20:22
but yes but cool yeah those can be really expensive
20:28
if anybody wants like uh you know to know how to do it the manual way we did a podcast last week which i'm sure
20:34
neither of you have watched i i just i i joined i was busy
20:39
i was at a show which was miserable we'll talk about a different story but where i was going to show with my kids
20:44
my wife it was miserable so i just logged in i didn't listen to you and i just [ __ ] posted and then left
20:51
so i think it was yeah it seemed good it went two and a half hours so it's interesting that when you're on the
20:56
podcast with us and we start talking after minute 90 you start hitting the button but if we're not there and it's just you
21:02
then you can run for two hours and 20 minutes and it's totally good the 90-minute mandate is
21:08
from the fray i would talk for three [ __ ] hours but so schultzy did not have admin rights to
21:15
cut you off right so it was like
21:20
that's why he's not here today is because he had enough last week he's already done two weeks of shows
21:26
with one he's like i've i've done my time lapse i've done my time that's right yeah he's like two weeks okay so i
21:32
see how it goes so when when we talk after minute 90 you you got a finger on the button but if we're not here and
21:38
you're on your own [ __ ] yep
21:44
i got rug the two weeks before that before my week off i got i was the one that got rugged after mid 90s so
21:50
you know just saying i fly out the chat several times i said
21:56
do you want me to hurry up the thing with having no people in a chat is that no one can tell you this
22:03
the [ __ ] off um because they all left out for 90 minutes
22:09
we have had a comment related to just that bit we were talking about before we just started to talk about how burned out we are basically um and how burnt
22:15
our callum is and then monkey dust because i'm really good at tangent um ben davis has said uh i'll film the screen
22:23
this is related to the the growth fund stuff i think a better approach would be for the chain to pass a brief for a sub-doubt and then people could apply to
22:30
be part of the sub-dial and then you know then the reports would allow people to
22:36
reapply or step aside yeah um so that's actually pretty that that's
22:41
probably not a bad shout out actually in terms of like you know crowd crowdsourcing some of that wisdom in terms of um bringing on additional
22:47
people i think for something like something like non-technical skills
22:53
that's is in really specifically non-engineering skills that's probably a pretty good shot actually
22:58
um i guess sorry i'm sorry catch up go ahead please
23:05
no i was just gonna say like because obviously you can do a lot if you don't properly vet an engineer they
23:10
they've got access to your source code you know if you actually allow them access to the repo they could i guess
23:16
potentially do more damage like but then they could just like fork your code submit a pr with a bug or a
23:21
vulnerability and if you miss it so maybe it's the same it's the same risk it's just trust isn't it
23:27
well i think i think it in that subdial area if you have a small team that's available to do that then you could you
23:33
could pick a start and end date and then you could stop it and you can measure the performance you could say we want to
23:38
be able to do elections on those members or something else right like you can get to a more of a
23:44
structure it becomes more like a board right you have a board associated to a vc that has performance metrics and
23:50
goals and then and then whatever i think i mean this is all stuff that will come as the numbers increase and it
23:55
has um as there's more failures i think in terms of funding and everything else
24:00
right now there's so it seems like there's so many tokens available that you could just kind of throw [ __ ] at whatever and nobody's going oh wow you
24:07
know there's a real problem here it's too early for that but after a couple years i'm sure those process will come in just because it has to
24:14
but at the same time i was having a conversation with some of the junior swap guys earlier where they were talking about um
24:20
you know the trouble of scaling in terms of hiring um and
24:25
i was sort of saying well you know if you would you rather be in the position in a few years where you've got
24:33
you know maybe other services have overtaken you or it it's clear that you've run to the end of your course
24:39
and you've still got you know zillions of tokens yeah or would you
24:44
rather be in the position where you still [ __ ] it but there's nothing in the bank you gave
24:50
it your best shot you know the community everybody who's invested in your success
24:56
looks at what you've done they go well didn't pan out the way we all thought the way we were hoped
25:02
but you know really did give it the best run spent money where it was necessary like is that it's that thing of like you know
25:08
there's there's two things there's like the effort everybody involved and then there's the tokens or the money or whatever right like you don't want to
25:14
end up in a position where you fail with leftover energy and leftover money do you sure yeah no i
25:20
it's totally right right and the other part of it is like we were talking about other chains everything else the other
25:26
part of it is too much cash like when you have too much cash that was that was what i think noah was talking about before is like you get
25:32
you want enough cash to actually pay people to eat like obviously that's there like we've talked to you know
25:37
you're talking about different changes people gotta be hungry a little bit hungry but not like not fat
25:44
right um but the other thing is like you have right now you have you know a16z or you
25:49
have the other ones that are throwing hundreds or 200 or 300 million dollars at small teams
25:54
and honestly they have the same issue they can't find the people right and so what that turns into is it
26:00
turns into an overstressed engineering team and a massive marketing team
26:05
right and that's not necessarily anything good either right um and honestly and a lot of things we were talking
26:10
about before that cosmos is a better situation with that was it i don't know if it was this or somewhere else we talked about actually
26:16
was on twitter i was talking about um somebody was making fun of the marketing of a different chain and and i
26:22
hopped in to say like cosmos marketing is awful like this is a this is a great technology
26:28
chain that i think is under marketed and it's and i think for the most part it's there's a lot of engineers saying oh it's going to work out it's going to
26:34
work out and best technology does not work out like that's not the way that life works
26:39
betamax right exactly there's there's a ton of those stories right
26:44
and best technology is not a guaranteed winner even with ibc and everything else like you have to have the right mix of
26:50
both the technology and the emotional side of it and if it's if it's not available and you don't have
26:57
the right marketing and you don't have the right visibility it could be [ __ ] awesome it's not gonna die so especially in this space right
27:04
yeah i actually wrote my i wrote my original dissertation on a version of that which is basically about how
27:11
um essentially academic theory and rev there were a bunch of people in the
27:16
early 20th century who wrote about how academic
27:22
theorists understated how change and especially revolution actually happens and how emotional the trigger is and how
27:28
irrational it can be but also how understated the impact of violence is
27:34
and how this was the bit that i kind of wrote about was like how people's reaction to that is highly based on like
27:39
it's not like some like rational world view of like here are the values i'll accept and hear the values i won't accept it's like
27:46
you know actually there are some things people will fight for like physically you know imply violence and go you know
27:51
i'm actually i'm actually fine with that and and all this sort of stuff and how people's essentially people who have a
27:56
real really deep technical understanding and spend a lot of time thinking about a thing whether that's social change whether that's
28:02
socialist revolution whether that's cosmos whether it's ibc tend to underestimate the tend to
28:09
misunderstand the way that people outside of that sphere are brought into the understanding and
28:15
emotional engagement with the thing that they want to see come to pass you know right yeah and i think somebody just said
28:21
[Music] hobby holder hey welcome uh said the market the market is network
28:27
effects and it is to a point and i agree with somebody else somebody else mentioned here around it
28:33
could be effect but there has to be a broader audience to be able to see it or won't do much i agree with that too like like network
28:38
effect has a piece of it something else but there there has to be more that like there has to be a greater visibility to it and we
28:44
talked a lot about just you know we've talked a lot about ui and like you know the lowering the barriers to
28:50
entry and understanding and clarity and those types of things and honestly you know um it'll it'll grow with the maturity of of
28:58
different types of solutions like liquid stake and other types of areas where i feel like that
29:03
that you don't have to understand the details of lockups and you're not understanding like there's a lot of things that you can kind of even though
29:09
it's more complicated from a technical perspective the actual ability for a human to understand what the hell's going on
29:15
i think somewhat you know makes it a little bit easier but so anyway i don't know how we get on
29:21
this topic i don't know where we came from this but uh yeah well i think we've we've been talking we've i guess we've been talking
29:26
about funding and growth haven't we so but and the other interesting things as well like like you were saying you said
29:32
where i think i think you're probably almost directly referencing sui however you pronounce it and an aptos which is
29:38
you know these lot these teams that are given a lot of money and have to scale which i mean to be fair i could also describe a bunch of teams of cosmos too
29:44
yeah um is that like some like apologies to why should be fair i
29:50
wouldn't dox them on the podcast anyway but a colleague of mine who's sadly i've forgotten which one a former
29:55
colleague of mine said that every in their experience every time a team scales by a factor of two
30:01
everything falls apart like whether that's from one person to two whether it's from two people to four
30:06
everything falls apart and like you have to completely restructure the way that you work together and like i guess that kind of makes sense like if you just
30:13
like i saw this with the name service like you when it's just one person you're just like twiddling away on your own and then right you bring in other
30:19
people to try and build something more significant it's like okay well like so the way we're like um calum and weaver who are the other two primary
30:26
devs on how like weaver does a lot of the front end but it also has quite a back end to be
30:31
fair callum does the lion's share of the rest the back end and actually like i haven't done as much code i've been
30:37
doing a lot of the organization and the legals and the like all the setup and all that kind of stuff yeah sure
30:44
but but it's that thing of like you can't assume that your role is going to say static either or like
30:51
how a team dynamic works out is always like it's very it's very very very hard
30:56
to work with other people it's very very hard to bring people in and be okay with how the dynamic works out and what that
31:02
will mean you have to do as well yeah you know because i'd rather be coding but it's not
31:08
over and over again like everything i do i always end up getting pushed into organization and management and that's the other thing like you see
31:15
you will see with core teams with founders like i've worked for three uh two or
31:20
three startups where i just spent years working there telling the cto to
31:25
stop just trying to skive off and code and actually manage the team and set a road
31:31
map and and like i understand it you know if i was in the situation i'd find it very frustrating too but it's
31:37
i don't know i'm not yeah it's just like scaling is super super hard and i think like
31:43
even if we had the engineers not just knocking around which we demonstrably don't
31:48
right we'd still have all that we'd still then have a second problem to deal with so
31:54
we've got like this huge problem actually bringing people in and then we've got like the second problem waiting which is then how we actually
31:59
build effective i don't know it's it's an interesting one like you wonder whether there's a
32:05
there's a good way past it and and i think also back to even just like our little teams here like
32:11
the majority of game of game of nodes plus game of nodes friends of the show we're all like
32:17
one person with freelancers or one person with maybe one other person helping or at the most two people right
32:23
almost all the validator operations like you sir rhinos too right
32:29
yup i think you might be the outlier i think you might you you've like oh you're overstaffed compared to the rest
32:34
of us well i think the other piece of that is you know not to relate our our issues
32:40
are quite easy because we're primarily infrastructure type security so it's pretty along the lines of what we have
32:45
but in your piece like going back to hal and with callum and i'm sorry what was the other name that you mentioned
32:51
uh we the waiver so that type of thing i could totally understand how that that doubling piece comes in and also with
32:58
the other thing i think is interesting here is the speed of this market like in a web 2 world if you three we're
33:05
building howell there's no expectations there's no kind of like timeline around it and
33:10
everything else and then here you know you guys get you you open up a public beta and that's instantly
33:16
you know around when's the drop happening and what's the airdrop and this and that when's this coming dude like the expectations i think are much
33:21
higher honestly in crypto and especially in the juno chain too because there's so many projects that are going on so that
33:26
means you guys have to gel faster and you have to determine roles faster and like i know it's a lot of outside
33:32
pressure that i think maybe doesn't exist when if you if you guys are just building this in a in a vacuum and a database you could take as long as you
33:38
want it right there's no big pressure here but here in this it's very much more visible i think project
33:43
infrastructure i don't agree with that or not but it sure seems like it kind of working in
33:48
both worlds it seems like yeah there's a lot more like expectation or at least the expectations higher
33:53
and like you know you go into how the beta and something doesn't work and everybody's up your ass because because
33:59
something's broken or something else it's a blessing and a curse isn't it so like you know i've equally worked for i've easily worked on things in the past
34:05
where we were we were killed for a thousand users right and literally killed for a thousand users
34:11
who told us what features they wanted you know we those companies were spending
34:16
money on people to go out and talk to prospective users or users or do surveys you know
34:22
and you know the the the and actually again this was a conversation i was having with some of the junior swap guys
34:27
actually the other day um was you know like in in crypto you have you have these people who will tell you
34:34
straight to your face i love your project but i hate that i
34:40
don't have x and y and z like and yeah i feel like that probably feels pretty bad if you have like thin
34:46
skin i know sometimes it it gets on my tits as well to be perfectly honest like when people are very very forthright but
34:52
then you take a step back and you go like this is what this is kind of awesome like it's better that people care about
34:58
it than not for sure at all um but not too much
35:04
not too much but maybe that's the benefit of maybe the validator thing i think the validated thing you're
35:09
until something like prop 16 happens uh you're quite invisible
35:14
but you're just kind of running at this constant high level of anxiety right but it's quite hidden until
35:21
suddenly you're in the limelight and people are sort of like have a very active opinion on what you do you know
35:26
yeah maybe that's where it's a bit different to like you know building a application being an application developer in the space let's say or
35:32
building a chain wait building chain probably is the most visible you could be i guess i think
35:38
that adds to your burnout too right because all that all that you just talked about add stress and it adds time and it adds thought
35:44
right and we're this is the burnout episode i'm trying to bring it back but i think that all adds to a piece of of
35:51
um that expectation creates you know it creates uh it creates uh weight right on the
35:58
shoulders would you yeah or no yeah
36:04
i guess i think it's more you know what i think it's more the i think it's more the
36:15
i think it's more that yeah when you have people the the flip side of the coin is the yeah the weight of
36:21
expectation means you have a lot of things to think about and each one of them means that you
36:26
if you focus on any of those things too much yes you're
36:31
going to then lose focus and not do one thing well and i think like
36:38
it depends i guess what you how you like to accomplish tasks but i think a lot of technical minded people like to
36:45
do a thing and do it well and then move on and be comfortable that they can sleep at night about that thing i think
36:51
that everybody has a sliding scale of perfectionism but like especially in infrastructure that's kind of the way i
36:57
feel about stuff and like you know you were talking about all the chain launches and stuff earlier and noah was talking about all of the test nets that
37:03
are running and stuff at the moment yeah and that's one of those where literally i it's not quite a panic attack but
37:09
because i i don't think i have an anxiety disorder but um but i i can feel like this like you
37:15
know like the high-pitched wine and i was like ah this is what people who have anxiety this would probably be like a
37:20
trigger or something for that yeah all those things you're like that's a lot of things to think about that's
37:26
makes me feel tired thinking about it and that that's so go ahead well no no
37:31
well i was just gonna say like in terms of yeah it's it's the workload like
37:36
there is so much going on one person just doesn't have enough bandwidth to stay
37:41
over the over everything so i mean you're sort of forced to be what what is the most interesting to you
37:47
etc like i think talking about team sizes and and you
37:53
know that type of stuff i think if if uh king nodes were going to get more people
37:58
then we would probably not even look for a technical person and probably more
38:03
someone like um like a marketing person uh
38:09
as well as probably i would like to have an analyst um
38:14
of to evaluate chains instead of a technical person someone who can go through and evaluate whether or not
38:22
they're like worthwhile software or whether they're just another [ __ ]
38:27
chain or something like that so i think technically was like with
38:32
you know i'm the main technical guy there's enough capacity just in myself to take care of
38:38
what we're on because we don't join every chain um but i also do everything manually i don't i don't really use automation
38:45
for booting up nodes and stuff like that because they can be done quite quickly anyway
38:51
um so yeah i mean i think
38:57
i think you just have to want to and and but you need to bring your scope of work
39:03
down so that you're not burning yourself out to the things that you're you're the best at i mean it's a little bit harder
39:10
for you the fray because you're working on multiple projects and
39:15
you are two things you're a coder and you're a node operator
39:21
so those two things this is not like there's a little bit of um
39:27
crossover there but mainly like when you're working on how you're working on something that's completely different to
39:32
your your validated business and it's really going to divide your attention you know what i mean and you've only got
39:38
so much bandwidth for different things in your head well technically i spend i spend more time on juno really than how
39:44
most of the time well that's another another project as well so you're like a main developer in in juno but
39:51
like i think of things so one one thing that i um
39:57
think of like in terms of your available bandwidth right um in
40:02
your head i'm talking about not on the internet so um here's an example right like when when
40:10
you don't have a lot of money okay you think about different things than someone who has more money
40:16
and you've only got so much space in your head an example of this is like
40:22
say you're you're shopping right you're walking down the aisle you're like earning
40:27
minimum wage um and you've got a couple of kids in tow you might be a single mother or whatever right so you
40:34
might um be doing it a bit pretty tough and what you're thinking about when
40:40
you're walking down that aisle is like you're adding things up in your head or on a calculator where you're
40:45
picking things off the the shelf and you're putting it in your cart and you're very concerned
40:50
with when you get to the end of your shop like if you can afford to pay for all those things that are in your cart right
40:58
take someone who's like a middle income or even a higher income earner
41:04
walking a couple of kids same situation but just at a higher
41:09
income level right and more secure in their in their financial position they're walking down the cart they're
41:16
just picking stuff off the shelf they like and they think they might want to eat right they're just picking it off and chucking it in meanwhile in their
41:21
head they're not thinking they're not adding up prices they're they're they're thinking about you know work
41:28
tomorrow like what what have i got to tell jane to do you know different things
41:33
and and it's more available bandwidth for them to be able to put towards their activities of like
41:39
earning money or or family or relationships or anything else right that
41:45
it's not they're thinking about different stuff and they're they're being able to apply that elsewhere right
41:51
so for us like when you have multiple things going on that's
41:58
different places you have to allocate your your like internal bandwidth so
42:04
you know is that necessarily a bad thing i guess it depends on how um
42:10
like uh good you are at those things like how ingrained they are in you like if doing
42:16
something is easy for you then maybe it doesn't use a lot of bandwidth if doing something is
42:22
difficult or new then that uses a lot of thought and bandwidth right or if you're under pressure on that
42:28
side of things then you're applying a lot of bandwidth to that and then the other things in your life suffer because you can't
42:35
um you know apply your thinking to those other things does that all make sense
42:42
yeah i i actually agree with you a lot on that that entire assessment there is
42:47
a there is a i believe a considerable body of of work that shows that people who are time poor
42:54
um are stressed out by the fact they have so little time and they make poor decisions about time
43:00
uh management and people who are like people who are um you know have a lower income or who are
43:07
you know financially poorer also can make wrong decisions about money when they're
43:13
stressed out just like fully focused on that situation because there's only so
43:18
when you're already at like the burnout point you start making bad
43:23
calls basically so there's there is like a bandwidth thing built into the way we think about problem space that's our
43:30
primary like preoccupation if you like so
43:35
i guess to that end like i manage my time based on priority right
43:42
so there's a list of stuff that gets added to all the time
43:47
and it's ranked by priority or the highest yield um to me in my life
43:53
uh or you know success factors right so if applying my time to this particular
44:00
thing is going to increase the success of that thing then that's what gets my time if it ranks higher than the other things
44:06
on the list so you know if there's that sounds like the what's the that awful hedge fund guy
44:12
the if it isn't making me money i drop it well it's sad but true right you're not
44:17
gonna it's you have to apply yourself to the most um
44:23
to the best yield like you wouldn't if you're investing you wouldn't pick the thing that's got the worst outcome
44:31
possibility and invest all your money in it would you because it just isn't worth it so you would
44:37
apply like you would invest in the thing that's you know potentially got a high growth
44:43
possibility um the newer thing like you know you wouldn't be
44:48
you wouldn't be investing your thing your your money in like i don't know um
44:54
inflatable i know what you're gonna say as well
45:02
you wouldn't be investing your money in flammable diapers right oh flammable these inflatables inflatables was going
45:09
to start with inflamed okay i had to think something else that's not exactly that's not true though because like the three of us sit
45:15
on discord and talk [ __ ] about cosmos validation other types of things and that's not there's no roi in that
45:21
excellent there is an roi in that but but it is time boxed as well like i don't know about you but i open discord five to ten
45:28
minutes at a time while i'm making a coffee and then i close it several hours really yep wow i've
45:34
noticed i i have i have telegram this thing is on 24 7. i get beeped in the middle of night
45:40
it's ridiculous it's it's literally turned into it's irc for 2022 but it's exactly the
45:48
same because i want to know what's going on and also the other thing with that is is that the fomo piece of it which is we
45:53
were talking about before the missing out piece of it is that in this space when something drops like hey there's a test net coming
45:59
or there's a validator form or something else literally 10 minutes can make the difference
46:05
right and that that's a miserable life because there's some things that like we've gotten into certain situations
46:11
where it's just been lucky because we've at the right place at the right time right and that's helped us be able to grow and not only be right we're a good
46:17
validator and we're good infrastructure and security everything else but but yeah there's also a there's a mix of luck in this business as well and that's
46:23
that's tough to not be you can't really sleep on this business which is tough yeah but i mean well i just like the the
46:30
the impact i think the other thing is like you know the impact on friends and family is pretty major like
46:36
the when we were doing the run-up to monitor um last
46:41
autumn um there was a period in that final two test nets where
46:46
i actually like had a long-standing thing where i was going to see some friends that i hadn't seen for two years because of the pandemic yeah and it all
46:53
went wrong we we crashed the test net we had a hard deadline because the core team had said on twitter that we were
47:00
going to launch on the 15th and just instead of seeing my friends i was sat on their sofa
47:06
working for two days basically sure and you know i think that is the thing as well like you know cyber attack i was
47:13
spending some time with my dad for the first time in a while the first one in in april and it was just like okay see ya
47:20
you know whatever so it's i think it's like the cumulative effect of maybe some of those things too
47:26
and i know that like i've talked with quite a lot of the other validators that are friends of the show and stuff and i think all of us
47:32
have a story of where something family or holiday or something just got taken out back behind the barn and shot
47:38
in the head you know uh but the better that's the that's part of the business right yeah you're right
47:44
because the other part of it is that it's when i'm when that's not happening and when i'm sleeping
47:49
i'm making money because there is there the blockchain is moving right and so when those things don't happen
47:56
um you know i've had a 30-year career in consulting other types of things where you have to be on site and you have to
48:02
be working to be able to make and this business is not like that where you can be able to build a an income
48:10
that's based off of good planning and and other types of things people get into that position and the
48:16
the the the caveat to that is that every once in a while you ruin your wife's vacation
48:23
so so there's up here hey we're on vacation because of this business so every once in a while i need
48:29
to ruin it through through a 2 a.m you know hey we're not going to this show because
48:35
um something something right whatever it is it needs to happen so i think there's no free lunch right
48:42
yeah yeah it is it isn't it yeah i mean this is the thing you can't
48:47
i think all i think all of us who've worked in other areas you know before are kind of like
48:53
yeah we're here because we want to be but it doesn't make it any less like i think for most of us we're coming up
48:59
to about a year or most of our first main nets right so it's that thing when you look back and you're like yeah i've been i've been
49:05
like on 24 hours a day basically for for a year now used to start to go
49:11
uh yeah that's pretty intense yeah but i mean you're doing you're doing both development and and operator which is
49:18
which is tough because some of those are technology slash uptime requirements and other ones are
49:25
people requirements where you're you know make sure that people are not stuck or they're not delayed because of you or
49:30
they have what they need to complete their job the people is the people yeah you're right the people is harder than the tech but then like you know i look i
49:37
look at you and like null and some of the other like folks like whisper whisper and co who are up on like
49:42
12 13 chains like yeah there's only so far automation can take you when you're a team of two people you
49:49
know polka tube wow i just i just dumped that
49:54
i don't think poker zest here though that's the problem polka is very very good and i'm like i'm not just automation at i don't
50:01
know what the hell that what you would call that but when i go to that website like how many chains he has i i feel like a freaking
50:08
loser yeah i yeah i just i i
50:14
i feel like whenever i look at polka's stuff that like poker is poker is living
50:19
the organized ordered life technically that i wish i was and i'm just
50:26
i'm just not there i'm just not like i i just don't have that that
50:31
focus that detail i i can't focus on one thing that long there's there's thirty three
50:38
he's an exceptional in for a guy and people should delegate to him he's like he should be much hotter than a mini
50:43
check there's 32 chains on this website right like this is this is the case study for ansible like this is how can i
50:51
take a pretty complicated like this is a high uptime high stress this is this is a
50:56
compliment don't get me wrong how do i take a business as high up time high stress with downsides associated to
51:04
bad uptime and use proper um
51:09
automation to be able to build at scale and this is a good case study for that right i mean there's
51:15
polka has like across the board like great uptime and and use of technology and you have 32 chains on
51:22
this list that's crazy yeah i also think as well like the i
51:28
don't know if you've read polka's blogs on like their reasoning behind why they do things in a certain way as well like
51:34
it's it's really well thought through like as in
51:40
like okay i know i always get [ __ ] and i always talk about getting [ __ ] for running on aws but my reasoning for running on aws is basically like
51:46
it's what i do for an enterprise client and generally hasn't failed me yeah that would be a very short blog post pokers
51:53
is actually a blog post with actual technical reasoning and justifications for
51:59
decisions and i'm like again this is that attention to detail and
52:04
thinking it through that like i just uh yeah no i think they're an incredibly slept on validator yeah
52:10
obviously like i say like you know i i'd started um actually terraforming some of our stuff
52:16
because that's the automation framework that i used most recently in my day job a couple of years ago sure um
52:23
and then i saw that that paul crude basically answered everything and i used to work with antibond like 2014 it was
52:28
like [ __ ] it right it's already done it's already done and i explicitly trust that whatever his polka do has done
52:33
is better structure than anything i would do and i literally just like forked it adapted the bits that i wanted to change
52:39
ran it it worked and i was i said i think i sent polka a message or maybe maybe it was
52:44
in a public discord channel i was just like [ __ ] hell your answer was really well organized thank you
52:50
because i needed to change a bunch of things where i was like my setup's just very quite different and
52:55
it was just like it was organized enough there was somebody you know it was like several thousand lines of ansible or whatever and i was able to find
53:01
the bit in the right places well structured it's got a readme that's really short as well but it was just
53:07
well structured enough i found the bits i was like this is just really good i just i would
53:12
none of none of my stuff is that well organized none of my bash script folder is that well i just i doubted it for
53:18
anyone just just to add more detail on null's comment to delegate to polka chu only if
53:23
he's below the three of us otherwise screw him
53:32
lavender five and and move him down the chain let's let's let's be honest here
53:38
and the one time i did on on a chain on kerberos like i i wasn't in the set and
53:43
poker gave me like poker and jaby both gave me a few tokens to get started
53:49
they're just they're just great guys like obviously you can't say you know japanese everyone in his jab is a great guy but like polka's such a great guy
53:55
yeah are you still on kerberos i am now well i want to tell you
54:02
so um just you know actually what you should do is
54:07
undelegate from everybody else and just delegate to
54:13
us in poker [Laughter] are you guys like top three or something
54:20
uh i'm actually i'm not right i'm talking about on every network man i'm number four kerberos i just actually
54:26
moved down because friends moved up but you know not friends of the show when kerberos launched i thought to
54:32
myself no thanks my my empire on kerberos as number four
54:39
validator equates to 1400 usd so there you go
54:44
that is savage raking it over there yeah but i hope
54:50
maybe i'll turn into something let's see but you know there's also a thing here you know when we were talking about um
54:57
you were talking like validators dumping like this is way back during like when when the price was still pretty good
55:03
yeah but like you know you do not get to a point now where you know there are chains building in the in the bear
55:08
there's a lot of stuff going on and like to kind of again back to the kind of keeping up with all of it like
55:14
if you were going to be looking at hiring looking at getting more support looking at doing more
55:19
ambitious stuff as a validator organization to take like i guess to take your business forward or to take
55:26
whatever your long-term plan is like you know whether that is like like us getting
55:32
going more into the kind of application side of things or you know whatever direction you want to go
55:37
that just again shows how important it is to have made sure that you've got a firm position
55:43
from when the market's up because right now like you wouldn't be expansion forget
55:48
about it like that's not happening in this market and yet the chain will just keep coming the noise level isn't going down the workload isn't going down and
55:55
that's it and like like you want to be able to participate in these changes like we i've talked to a bunch of great
56:00
startups and and groups that are coming um skip i just talked to last night like that team sounds awesome we've had
56:07
conversations with d fund like they're you know obviously kanto just went live a little bit ago and then empower talked
56:13
about today like you get in a situation where like this goes back to what we talked about
56:18
for on the dow side of the house is like you want to be able to do due diligence but as an individual you need
56:24
to how much time can you spend towards that and how much time can you actually understand if of what's there and so i usually do team
56:31
kind of uh reviews to see what's going on but there's a huge fomo piece within this and honestly this is just starting
56:38
right like we're at the beginning of this ramp in terms of the number of chains they're going to launch and everything
56:43
else so if we're looking if we're looking this tired right now it's going to be bad in six months from now
56:49
it's definitely yeah you can't be everywhere you can't be everywhere not be everywhere at once unless you've got
56:54
unless you're like yeah but also i mean there's two there's different strokes for different folks right so there's yeah there's um pokechu
57:01
does i think it's just him but he does everything like automated in a you know
57:07
like you said you know um i think the pokemon intern account is just poker's old isn't it it's just yeah
57:14
it's it's at nighttime don't dox the guy what are you doing ah well
57:20
the phrase did it anyway well i mean that's just my that's just my suspicion i can't prove it but
57:27
if you look at then um cosmo station cosmos station has a shitload of people and money right
57:33
yeah so i mean they're both on every chain one yeah
57:40
like i was talking to some of those guys in um [ __ ] where was that place that i went
57:46
uh prague that one the thing the cosmos event yeah
57:52
and and yeah i was just like oh just out of curiosity how many staff do you have there's some of these guys like 50 and i'm like
57:57
what and now i'm like how many engineers and they're like oh yeah only about 12 and you're like
58:03
like heavy breathing you're like [Laughter]
58:10
um so just on the automation thing uh dan from
58:16
defiant nodes i'm pretty sure it is do you know do you know of dan um he's doing some work with um
58:24
strangelove and i've been following along with him on
58:29
discord and he sent me um
58:34
a little thing that he's been working on the other day with strange love with the whole
58:41
k9s or k8s um kubernetes
58:46
yes so he's sent me like it's in a cli a dashboard
58:52
of i'm pretty sure he's got automated launching everything into k8s and it
58:58
looks pretty cool to be honest and i want him to give me a run through and i think he's going to give me a run
59:04
through today at some point that's a patient extension of what's going on with the uh the docker structure
59:11
yeah like the test i think they're running k8s running everything through
59:16
docker um so the example he sent me is for kujira and in here
59:24
um on this k8 he has got everything running automated with uh cosmos
59:29
exporter graphana centuries node exporter
59:35
price feeder and prometheus all sort of automated to mesh together yeah
59:41
so it'd be interesting to see it looks quite nice i i want to see um
59:47
the scripts and how quick it boots up but uh yeah people are people
59:52
who work on ultimate solutions to like the automation so
59:58
i'm sure what was a what we've talked about before what's a container structure that's that was built for originally for akash to be
1:00:04
able to run nodes and akash um oh so that's um the tom tom one that's uh omnibus
1:00:12
so it's basically an extension of omnibus right you're basically just orchestrating containers for for those types of things i mean it
1:00:18
could be or it could be their own custom containers i'm not 100 sure right right i guess
1:00:23
they're probably i would imagine if it's be the same containers because they'll
1:00:28
just be chained by chains so they'll just grab them off of the they'll just grab them out of a known
1:00:35
stable repository right sure um so kubernetes is just the
1:00:40
uh to those listening who might not know what what the [ __ ] we're talking about uh docker's a virtualization environment
1:00:47
so you can run run a thing in a docker container and if your system can run the docker container
1:00:52
there's a higher likelihood like whatever's inside is kind of isolated from the host ish a bit of hand waving if you can run
1:00:58
docker um so it's a way of like running a thing on multiple devices as if it was
1:01:04
a binary like just an executable program and kubernetes is an abstraction
1:01:10
framework for running docker containers um so you make a kubernetes cluster
1:01:16
across one or multiple machines and then that cluster just
1:01:21
has capacity to accept things deployed to it network them together and offer
1:01:26
ingress and egress of data to them essentially via whatever scheme
1:01:32
you want so kubernetes and helm are two of the main the helm is the way you kind of
1:01:37
define large deployments of kubernetes stuff generally speaking
1:01:42
um and the connection here is the akash overclock did a lot of work on
1:01:48
kubernetes and helm which is where they're coming from with their software their sdl their software
1:01:53
description language is like quite familiar very similar to yeah kind of like a yamali version of uh
1:02:01
the terraform that you'd use um and obviously a lot of the helm and kubernetes stuff is yaml so that's why
1:02:07
the sdl looks a little bit like that's because um akasha's team have prior in that area
1:02:13
and that's why providers if you kind of wave your hand the cache concept of a provider is kind of like
1:02:20
a single or multi-machine cluster if you like there's a lot of the same kind of concepts
1:02:25
um going on there but like this is relevant because so the thing that we were talking about earlier which is that the
1:02:32
the aptos team give you the terraform for what they consider a baseline
1:02:38
implementation and you can obviously change it up if you need to but like so we're for this current test now experimenting with just running
1:02:44
straight up on aws um a kubernetes cluster um on
1:02:50
we're using hosted kubernetes so we're actually using aks aks eks
1:02:56
um but you could equally spit up a host and then throw create the kubernetes cluster and then
1:03:02
throw part of that terraform at the actual creation of the docker containers in the cluster right um
1:03:09
if you wanted and so tom from ecostake who's obviously a maintainer of omnibus
1:03:15
he as i understand it uses uh like he uses his docker stuff but he obviously he
1:03:21
just throws he has basically has a host that can take the docking containers he can run them i don't think it's kubernetes that he runs on
1:03:27
um but it's like a sliding scale from basically like i have a host that will run docker containers that i throw at it
1:03:32
all the way to i have several hosts that are essentially a kubernetes cluster i throw stuff at them it gets run
1:03:39
um and all this is just to be able to build repeatability in terms of
1:03:44
deliverability so you make less mistakes where it's very easy to make mistakes and
1:03:50
if you're expanding out from a you know from a bare metal perspective other types of things you forget something or something somewhere that you have a
1:03:56
playbook to be able to run you know those types of things the idea is that you're building automation to be able to
1:04:01
continually improve that playbook to be able to reduce mistakes and those types of things i don't necessarily
1:04:06
that's really what it's about it's not necessarily going to guarantee you know for people listening it's not going to say guarantee higher up time or other
1:04:12
types of things it's really um it can in certain situations but for
1:04:18
the most part yeah it's primarily maybe in a disaster recovery situation yeah you're right but but you know even even
1:04:24
in that case like you know the the scale of some of these deployments
1:04:30
like the provisioning and deploying the the aptos stuff you're waiting half an hour for it to
1:04:37
provision all the stuff the first time right so if you had to fully disaster recover that onto another cloud provider
1:04:44
or another um or if you had to fail over to bare metal if you hadn't already prepared a
1:04:50
kubernetes cluster ready to accept a deployment bearing in mind the fact you'd have to have separate maintained
1:04:56
terraform for the bit you wanted to do which is just the deployment which you would have had to have tested
1:05:02
taking time and cost and then maintained over time taking time and cost and energy that's
1:05:07
right yeah it still wouldn't be instant and this is why there's a sliding scale for automation right whereas something
1:05:14
something like what polka's done very very sensible right because i can spin up a really cheap hex in a box i can run
1:05:19
the ansible against it and see if it provisions of course well i actually don't even have it provisioning cosmos nodes or anything because i prefer to do
1:05:25
that bit manually um because we use tmk-ms and i prefer to just make sure i've switched all the
1:05:31
keys over and everything but i like everything to be i like it to install prometheus node exporter all
1:05:37
that stuff not turn it on but like have it ready on the box so that bit i can just throw up against the
1:05:43
cheapest hats in the box and just check it works yep you know and to me that's the thing that like every now i guess
1:05:49
going forward probably every few weeks i'll just run it against a blank box and then throw the box away just to
1:05:54
check it's ready to go and that's there's downsides too we just talked about if it happened uh i think last
1:05:59
week maybe when nell and schulze were talking for six hours on this podcast but um
1:06:05
block damon double signed on evmos
1:06:10
is that what it was was it devos i forget where they double signed oh yeah there's two that double signed evmus
1:06:16
well they they white label somebody else right so they they were i think the understanding was that they were they had they're heavily automated right like
1:06:23
blockchain a huge validator like like back to your 50 person type of thing right
1:06:28
the downside of automation sometimes is in the wrong hands or with forgetting
1:06:33
what exactly that does or kind of understanding the scope of that automation you get to situations where
1:06:38
you make massive mistakes because you're letting scripts run and do things that but what freya was just talking about is
1:06:44
you know he's he's stopping to think about what the role of tmk ms is in that situation to make sure that i have you
1:06:51
know this type of structure where i'm not going to double sign blah blah blah or he understands the blockchain um and they just double signed on both
1:06:57
their node as well as a white label node because they were running automations and so there's a there's a there's
1:07:04
you know that there's double i guess because there's two sides of that sword right yeah i think yeah when you're talking
1:07:11
about signing like you either have to have robust built-in
1:07:18
um you know uh checking in your scripts or you need to just do
1:07:25
it manually because i think the only safe way to do it is just to do it manually no matter how much checking you
1:07:31
do in your scripting there's always the edge case and the the risk
1:07:36
well i mean the the outcome from that edge case is catastrophic so it's not worth the risk
1:07:43
i think yeah i don't think i would ever have automation turn on a node for me for example even
1:07:49
maybe an rpc definitely not a node even if i was like you never turn on a node you just have
1:07:54
to have it be a dumb disposable node and then you just manually point your scientists to it
1:07:59
um yes that's the thing i wouldn't even i wouldn't even turn on an rpc that was
1:08:06
open to a signer without double checking the firewall double
1:08:12
checking i'd removed the yeah the keys were not the ones i thought
1:08:18
all of that like i it's just not worth it like yeah um if i was doing rpcs
1:08:27
with like ansible or something like that i would probably have a script to set up um the node and the firewall and i would
1:08:35
run everything through reverse proxy and then just check the um caddy file or something
1:08:41
like that so i mean opening firewall ports is not a
1:08:47
massive drama if you don't have to do it all the time so as long as you run it all through your ssh port or something
1:08:52
like that and you have appropriate scripts that just adds to a caddy file
1:08:59
and restarts caddy i don't think there's a high risk or like a large footprint for it to
1:09:07
to um add security issues so yeah i'm not really worried about security issues so much as double signing like yeah i
1:09:14
mean reverse yeah for i would run a reverse proxy in front of an rpc but just you know for
1:09:20
for all the purposes i think you just opened up open a port don't you um
1:09:26
but uh yeah i was gonna so i noticed we had some
1:09:31
comments as well i just wanted to pull out a couple of uh so um
1:09:36
rama had said a little bit earlier there's lots to talk about difficulties on boarding people in multiple areas is
1:09:41
there a plan to address it in a decentralized world there is no one to take ownership
1:09:46
this is yeah it's it's pretty hard like the the approach
1:09:53
we're taking um like as in our envoys needlecast i think is we're just going to proactively hire we're just going to
1:09:59
hire and train and just we're just gonna have to try and cover that salary
1:10:05
by taking them on our engineering work um because
1:10:10
it's chicken and egg isn't it like especially with especially with core dev and stuff like that where like what i've
1:10:16
seen with causing wasm dev is that causing wasm dev is very very easy to sell people there's a lot of people who want to write rust there's a lot of
1:10:22
people who understand the smart contract thing there's a lot of people who for whatever reason see it as lower barrier to entry than
1:10:29
chain dev um i think i'm not the only person i know there is a movement about saying that
1:10:35
lean cosmos where it's mainly just a base chain plus cosmism is likely gonna be the one the dominant
1:10:41
paradigms for cosmos moving forward um i mean i personally go even further
1:10:46
and say actually just just ibc and then don't assume anything else about the stack is probably going to be
1:10:52
the dominant paradigm eventually maybe cosmos will be part of that probably it will be um
1:10:58
but it's interesting that that is it's it's very easy it seems to sell the smart contracts and rust
1:11:04
it's much harder and this is because the sdk is inherently more complicated to work with
1:11:09
but i think i think go does play a role and i've kind of i've beaten that horse so
1:11:15
um they should have a really light version of the sdk like you say but i don't
1:11:21
think it should include cosmoism because there's a lot of attack surface on cosmos oh yeah no but the so the way it
1:11:27
is implemented in the so there isn't a rus sdk but the way ibc is implemented is it's just a library it's a module
1:11:34
it's not a module module it's literally a rust library you can include it in any rust
1:11:39
project yeah you have ibc but it won't obviously work with all the other without all the
1:11:44
other stuff so it's but it's up to you how you actually integrate that into your application so that's why i think like
1:11:50
if you had cosmasms in essentially the bits the x wasm that's
1:11:56
in wasmd as a similar rust module then you would have ibc cosmosim
1:12:02
you can build into any rust project and you already have 10 minute rust
1:12:08
so i mean that would be my bet for the future is like a a much better way for onboarding to to
1:12:15
answer to answer your question rama um i like i know larry who's i think mars
1:12:21
protocol has also been saying a similar thing on twitter
1:12:26
and i think they're also he's a very smart guy but the confirmation bias of course
1:12:34
and actually rama made a second point which is just saying i've seen the community openly [ __ ] on people for trying to get into roles or positions to
1:12:40
assist juno as an example so i guess this is a this is a reply to that comment we had earlier saying
1:12:46
you know should we make a doubt and then invite people to apply right
1:12:51
so there are there are pitfalls maybe to that approach i suppose um
1:12:57
i guess like you know with the sub-diaz maybe there should be more of a
1:13:04
effort to look to the broader community for people for those sub-days who are
1:13:09
interested rather than um you know people that are already tied to the community might be a good way to
1:13:16
to bring in additional like network effects again you bring in a couple of people they've got friends who have got
1:13:21
friends who've got friends so um then you know you can't get more interested than if you're on a dow
1:13:28
and and focused on that so yeah i think it's good points there's so much vc money flowing around
1:13:34
now i think it's a good balance of that too like i mean we talked about four round you know what the role that bc is
1:13:39
and the experience there and those types of things and quote-unquote web 2 and all this type of [ __ ] but uh there's value in that
1:13:46
and there's value in in trying to separate you know teams and and
1:13:53
approaches to figure out what the market will accept and there's also value in
1:13:58
and maybe you know juno subdials taking larger chances because they have the
1:14:03
coins to do it so i do think that there should be some accountability though like there that idea around um you know
1:14:10
kind of building that dow team and then understanding over a course of six months what do they invest in and where those
1:14:16
teams are and if that makes sense or not and trying to make the sense of like are we getting an roi because
1:14:21
there should be no matter what there should be an roi at a portfolio level maybe not an
1:14:26
investment level but at a portfolio level i should look back and say hey we're making the right decisions here like you know going back to our point
1:14:32
before we're on marketing and you know the growth of the cosmos and those types of things like that it's important to
1:14:37
be able to put money down and be able to see that we're seeing return off that so um i don't know it's a good balance
1:14:44
between you know kind of the obviously the juno dao and also what's happening you know externally and and uh there's a lot of change starting
1:14:50
from a good amount of vc funding and there's still there's dollars coming in the cosmos which is good stuff
1:14:58
so completely unrelated um yeah my mouse stopped working too so i kind
1:15:04
of click around but um so the earlier you were talking about
1:15:09
your your dark british comedy comedy uh animation and um
1:15:16
just the only thing i could think about what you're talking about was uh douglas denim
1:15:26
you mean chris chris morris's character or do you mean um yeah i think it's chris morris uh not not the old denim
1:15:32
the young denim oh yeah um uh what's his face who's who's toast of
1:15:38
london who's in um he's in gothenburg's dark place right richard i don't know no no no no no
1:15:43
richard um um matt harry matt berry perry yeah yeah
1:15:50
he's just so funny man i love that um
1:15:55
you computer man fix my pants [Laughter] i'll hand me priest um yeah great um
1:16:04
even the old guy he's just like so so business uh he's he's actually he features very
1:16:10
strongly in a very good video by the band three track tigers they literally just so
1:16:15
okay two facts about mike barry number one he's in the video reset by three shot tigers they literally just i think they sent him a message like matt we
1:16:21
have this insane idea for a video would you be doing it he's just like yeah [ __ ] it what not um
1:16:26
number two he's in a band called matt berry and the maples because he is a folk singer and people people go to his
1:16:33
shows assuming that it's a joke and he is no he's it's it's a serious dead serious not a joke folk band he plays at
1:16:40
folk festivals and then some people go along and they're like oh right i was expecting musical comedy
1:16:46
no it's just a folk band playing folk standards it's like like like steve martin played the banjo you're like ah
1:16:52
it's funny then you're like oh [ __ ] this guy's a really good banjo player like where the hell that come from
1:16:57
yeah he there's also like especially from the period like going way back like 15 or so years or so early 2000s he
1:17:04
whenever there was like because he can i think play quite a few instruments whenever there was like a random bbc sketch show where they just needed like
1:17:10
a bit player to play like an old 70s prog guy or something he always like
1:17:16
there's always lots of young matt berry just in like shades and a mustache or something playing one of these roles but
1:17:21
it's just like there's this amazing one i can't remember what it is i think richard might be in the sketch as well
1:17:26
i'll try and find it for the show notes where there's he he basically he sat to this piano and they're talking to him he's like
1:17:32
supposed to be i was all prog rock guys from the 70s and they're just like what's the song about he talks about it's like really dark it's like about
1:17:37
his brother's suicide or something and they're like and they're like okay well um it's like it's kind of like the old
1:17:42
gray whistle test if you've seen that show until then i think it's richard ryder potentially he turns the camera like so we're going
1:17:48
to hear some of this it might be a little bit disturbing a bit moving but it's it's obviously a very beautiful piece take it away and like you sort of
1:17:56
he obviously starts like playing this sad piano and then he just sort of like kicks the
1:18:01
piano over and he's got like a keytar and they just starts like shredding and it's just like nonsense 70 synth and
1:18:07
he's just like rips his shirt gets his chest out and stuff and it's supposed to be the song
1:18:13
about like his dead brother or whatever i was like that is 70s prague that is exactly you've nailed it um but he wrote
1:18:19
like all those things i think he basically were they were just like we need a musical skit here's our idea and then they just got
1:18:24
to go write the music and then be in it so
1:18:30
mute i've said my thing now i'm using the thing matt barry mute
1:18:37
um i just looked up matt barry in the
1:18:44
maples and yeah man he's like playing in front of legit crowds like yes
1:18:49
that's the real thing like it's actually his side it's completely dead serious um he just looks so [ __ ] cool too like i think
1:18:56
he's probably quite a nice guy like i suspect he kind of seems like he's got his [ __ ] sorted um he's got the coolest
1:19:02
demeanor he's just like [Laughter]
1:19:07
you know he's like his background i think he is actually like a voice over artist that's like his thing
1:19:12
like so periodically you'll hear something on the radio in the uk and it's just matt berry like as in
1:19:18
before he was more well known you would just hear like a radio advert when you're driving along in like your parents car in the early 2000's or
1:19:24
whatever you'd be like is that that berry do you have another car buy a new one
1:19:30
exactly it would be like have you had an accident in the last and you're like it's matt barry it's definitely this is definitely yeah yeah and then like but
1:19:37
like now he's like considered more famous as like a comic actor and stuff but he still does i'm pr i'm pretty certain either or he's got impersonators
1:19:44
now i don't know it could be one or the other but there's a there's there is a he did
1:19:49
a more recent show called toast of london if you've not seen it which is all about he plays a struggling actor who's
1:19:55
basically the most awful person in the world and who this guy is also a voice actor
1:20:00
um but but is basically a failed stage actor at the end of his stage career struggling to make ends meet and it's
1:20:07
sort of like a comedy dark version of his own life where the mate where he's made like a version of himself if he was
1:20:13
just a prick to everyone all the time it's quite funny um and um
1:20:20
yeah there's a guy there's a few people in it who are like minor characters who have gone on to be quite famous in in the standard style of all
1:20:26
british comedy it seems where like the guy playing the t-boy in the studio is now famous because he was in that star
1:20:32
trek discovery and you're like wait that's the guy that was from the main characters of star trek discovery
1:20:37
and he's playing the t-boy that's always rude um well like ben we're sure ben does any
1:20:42
of you see um oh god what's it called nathan barley that's a really old one so
1:20:49
in nathan barley it's about startup startup culture in like brick lane in short like shoreditch east london in
1:20:55
like the early 2000s and it it hasn't aged a day it's still perfect is that the show name or is that the actor
1:21:02
nathan bali's name of the show it's about this guy who runs a startup which does nothing but is well funded
1:21:08
and they just do [ __ ] nothing and they're just all hipsters and everybody's horrible in it and
1:21:14
they do and it's it's so pitch perfect as a satire um but in it they have one
1:21:20
of their employees who's paid to actually do all the work i think he's the programmer that actually keeps the lights on yeah is ben wishaw who went on
1:21:27
to be like queue and stuff in james bond so he's like one of the biggest actors like most these other people did like one or two other comedy shows after this
1:21:33
and not much else and in it yeah so ben wisher is just like this guy who's in every episode and
1:21:38
they just prank him relentlessly and you're like that's ben we're sure he's the guy he's he's like his big hollywood guy now and
1:21:46
um and the guy who wrote it charlie brooker is the guy that created black mirror oh really oh yeah yeah so he was a
1:21:52
comedy writer he did a bunch of comedy shows like chris morris and people like that back in the day and obviously chris
1:21:57
morris is like chris morris wrote like day to day and stuff along with like commando and ucc and all those guys who did
1:22:03
alan partridge and nathan brothers from 2005. so that's like earlier yeah yeah yeah what what ever happened to the
1:22:10
mighty bush guy uh the mighty beach they did three series didn't they and then they they stopped
1:22:16
doing it yeah they stopped doing it it was a good show i think they ran out of steam london and i got
1:22:22
rack of toast at my local cafe [Laughter]
1:22:27
uh yeah we're gonna have to like so the the show notes for this is for this week is just gonna be like loads of
1:22:34
loads of like it's comedy somehow somehow there's always 90
1:22:40
minutes of content related to the cosmos even when there's no content well it's back to null's reasoning that
1:22:46
we basically have a voice call every week so we may as well film the voice call that's a podcast i mean that is the
1:22:52
whole purpose of this show just to record it that's fine provide entertainment um
1:22:58
coin water said something interesting there which is double signing should be addressed in the core code it is on my list to build but have to launch our
1:23:05
chain first we're running three parallel nodes that represents a single node failover and up
1:23:10
time is paramount we should not have to worry about slashing when pushing to do what's best for the chain
1:23:15
uh that's interesting but i guess the
1:23:21
the question is isn't like is there you don't want to guard against
1:23:26
double signing because double signing double signing is the attempt is potentially the attempt to interrupt
1:23:32
consensus with invalid state tears i think it's just so if somebody's come
1:23:39
up with well although actually god sorry to mention but i think it doesn't it works
1:23:44
slightly different in aptos doesn't it the the old double-sided roof because it's all threshold actually it's all
1:23:50
threshold signing isn't it hot stuff i've been watching a lot of videos around hot stuff and trying to
1:23:55
understand the differences between that and tendermint and kind of the benefits and whatever i but i'm double signing is
1:24:01
it's called it's called hot stuff yeah yeah stuff hot stuff hot stuff there's a good
1:24:06
consensus engine yeah that's something in the show notes yeah i found a good video on uh a16z
1:24:14
discussion around that versus tendermint and kind of the difference did you find it or did polka cheese say you're right
1:24:19
my bad my bad credit where credit is due as most things are the smart cast
1:24:25
this this podcast is just like we're just telling the things that polka do has told
1:24:33
i did find the second part of that video does that count yeah all right um
1:24:40
poco he hasn't been on before
1:24:47
um we were talked on a previous one of these with when schultzy was in about
1:24:52
the tim rough garden lectures about economic incentives which um is a talk course in the i think
1:24:58
optional course in computer science that he teaches wherever he is columbia um
1:25:04
and the online course content became really popular particularly with people in the crypto space he's started doing a
1:25:09
second series which is all about consensus engines and different consensus schemes um and his first deep dive series is in
1:25:18
on tendermint oh yeah so um that's another interesting recommend
1:25:23
it's [ __ ] long i mean it is courseware so it is here is a session with a virtual whiteboard i'm going to
1:25:29
tell you how the partial bft yeah
1:25:35
asynchronous partial partial uh what's it called bft works um so it is heavy going um but it is
1:25:41
interesting is it publicly available yeah yeah it's on youtube uh
1:25:47
i can send you a link um just go in my crazy youtube history which you can imagine is absolutely
1:25:53
madness yeah mine's mine's full of it's garbage i might just mind just matt barry and consensus algorithm
1:25:59
is that mine's like 8 000 one hour comedy
1:26:05
specials patrice o'neal and everything else yeah exactly there's a there's an amazing have you
1:26:10
ever seen that um instant regret youtube playlist that's one of the best things on the internet
1:26:15
where it's just all these 10 second clips of completely incomprehensible memes that have been gradually added to
1:26:21
over time and it's like it's like archaeology some of it because if you watch for more than
1:26:27
like a few minutes you'll be into memes from like six months ago and you'll be like
1:26:32
whatever cultural event this reference is i have no idea what's going on i did i did search and i
1:26:39
got the instant regret playlist one through 35 and that ended in 2021 so i'm
1:26:44
guessing we're up to the 50s yeah you're looking for there's like i think it's like instant regret playlist
1:26:51
instant regret clicking this playlist something like that it's by epic donut dude
1:26:57
on youtube um but we we once had a house party where we just put that we did body music we
1:27:04
just put on that like on the sound system with the tv and people kept walking into that room being like it's
1:27:10
perfect and then you come back 30 minutes later and they just stood there with their beer laughing and you'd be like well the party's over for them like
1:27:19
is it was it just a regret or mystic karma which one was it trying to find it
1:27:26
who's the who's the user i think it was epic no epic donut dude well epic donut dude
1:27:33
yeah obviously how is this i mean how could you you know how would you even why would you
1:27:39
accept any substitutes unless unless epic donut dude gave up the mantle um i have to admit
1:27:44
i've not i've not watched it in a few years so 113 000 subscribers far more than game of notes so i mean i'm
1:27:50
subscribed click i'm gonna watch the [ __ ] out of that
1:27:58
i had an idea for a bar 10 years ago i want to open a bar that would just have on the walls like a thousand five inch
1:28:04
screens that would only play like these types of videos like over and over like each screen would have its own meme over
1:28:10
and over and over and you have thousands of them there used to be a bar that he used to go to that would just play um archer
1:28:18
that's it it's archer or whatever just had archer going in the back like with subs it was just perfect oh in the
1:28:24
background they had like you know i don't know metal and then just archer playing on all the screens
1:28:30
it was weird it sounds kind of cool yeah but also the first time i saw archer and then went
1:28:36
and looked it up and watched it all and it's great
1:28:43
um yeah anyway so i mean uh kevin here has a has a nice
1:28:49
comment i love when nell gets caddy i don't know what that is in relation to but
1:28:54
yeah every episode i get caddy yeah maybe
1:29:01
please never do that again but yeah uh kevin obviously famously of
1:29:09
the the tdf um see kevin's the marketing guy that i need um
1:29:15
i think i think kevin can do it i think kevin's marketing is probably the best out of any of us isn't it by far oh yeah
1:29:21
well by far with the pitches
1:29:28
definitely great marketing if i had uh a kevin for a marketing guy
1:29:34
um and a help tech guy that'd be great and if i had a rama for a research guy that would
1:29:40
also be great if you're a club or a rama reach out maybe we can talk
1:29:45
i need a kevin a rama and a fray for a developer that was that you're
1:29:58
you don't need to respond to every single tweet
1:30:08
foreign