0:00
by the way how okay well we're live
0:11
[Music]
0:19
hello and welcome to game of nodes a weekly podcast on the cosmos from independent validator teams and as you
0:25
probably guessed from that intro nel got set up but spotted it before it was too well too late too late
0:31
to say the thing that you're probably already going to say uh with us this week we have ethan frey the
0:37
uh what would you say founder creator uh private father
0:51
we're going to hopefully not expose our ignorance too much well we'll talk about some of the the motivation behind that
0:57
and some of the broader uh sort of cosmos topics because ethan also is a bit of an og in the cosmos space as well
1:05
um so i'm sure there are any number of topics and null
1:10
there are some topics in the spreadsheet in case you get lost so if you have any questions for ethan drop
1:17
them in the chat and we'll uh we'll pick them up as we go along as well as we as we find spots for them so
1:22
uh yeah that's pretty good thank you so anyway yeah thanks thanks very much for joining us ethan um i should we'll kick
1:29
it straight off with like probably the the most obvious question which is like how did you actually
1:35
get started in crypto and in the first place like what was your what was your in
1:42
uh all right i'll take the short story here my in was basically
1:47
i was uh nasa turned on early crypto that was cool tech
1:52
but when someone was telling me he had forgotten money to get any rules you wanted and you could make money like
1:58
it was basically governance it was basically a community government issued money that's how he made it so you know rather the bankers making money you
2:04
could have like you know whoever runs a child care makes expensive money and stuff like that and all kinds of
2:11
crazy rules of money like wow that's crazy money's program that concept just like read for me
2:17
you can totally reinvent society because like you know bitcoin is cool it's electronic money but like really just changing the rules
2:22
of money and being money this malleable programmable material um that's what blew my mind that's 2014. that's before
2:29
ethereum right this stuff was not possible but i got the idea and just kind of led me to find out tendermint and kind
2:35
of gotten the cosmos because i thought that was like a way of building this try to build a stuff in 2016-17
2:43
right and you were you actually worked it you worked at tendermint right back in the day as well
2:49
definitely i was like besides jay and ethan buckman i was a first dev working there
2:54
there's some marking people before that peng was there also early on designing brian started with me at the same time
3:01
january 2017. it was a interesting ride suddenly i'd like went from a normal job to being
3:06
paid with like you know i was paying a bitcoin because you have a bank account i'm like what huh
3:12
how did you find how did you find tendermint like back then like what was what what kind of jury to it and and how did
3:18
you even get involved initially so i mean the guy that read from me this guy uh
3:23
daniel who was pointing anyway go on that one he was trying to imagine
3:30
something way beyond him and actually kind of beyond blockchain even post-watching something like a holochain maybe even but uh he's throwing me stuff
3:37
he could throw me links to various projects and one of the theory of my little interesting other one that got my eye was aries industries
3:43
so i don't know if anyone knew a really old school here but in 2015 there's something called errors industries where
3:49
ethan buckley used to work um it turned into monarchs later on and i was trying to make legal contracts and that was
3:54
ethan buckman was and they were using like tenement 0.1 or something and ethan buckman later left them in
4:00
2016 to join jae kwon on tenement actually build tenement out and so i was like looking at that stuff
4:06
and then i was trying to figure out they taught myself go to get involved in it but that time time was their own product and um
4:12
yeah yeah started messing around with it it just seemed the right way of doing it
4:17
i tried to write solidity i look slitty i'm like this is impossible how do you test it how do you know it's working there's no bugs and there's like the
4:23
bugs glow everywhere right like how do you build solid testable software with this and uh then it looked at this like oh
4:28
it's go i can write solid tesla software and go like that actually like makes sense i would actually trust me something
4:34
complex there i think that's proven it was a little slow i thought the cosmos decay would launch a little
4:40
faster it did but yeah i definitely the idea of um of actually building solid software
4:46
these complex systems you want to build you cannot build those really so hard too complex
4:52
so um i have a question so so starting out right so you've decided that um you know you want to start
4:58
working on this project like it seems like a pretty big task to undertake so
5:05
um like how did you get started how did you plan it and how did you get started in it and did you put together a team or
5:12
did you just start hacking away on it yourself meaning cosmos or cause of muscle sorry because it wasn't yeah yeah uh because
5:18
it wasn't really a hackathon so we got a bunch of people who came together there's like five ctos came together around some
5:24
freaking beers so the [ __ ] would you do let's build bottom contracts that's fun and we did i didn't sleep at all
5:30
i i closed my eyes for two hours maybe an hour lots and lots of cool months and uh
5:36
and somehow ended up with this working i buy into sunday night it's an evening
5:41
i just have time yeah yeah it was pretty crazy and so after that i'm like dude they say grant's the winning team dev
5:47
grant to the winning team right to work on stuff so i you know went and applied the grant said hey guys let's all work out part time we got the jobs let's work
5:53
on part time and uh no one else wanted to really they also they're too busy so i said [ __ ] it i'll just quit my job and
5:59
work on this it's awesome i don't know i kept contracting getting aside for regen for
6:05
the first like few months i was like doing that part time and doing regen part time um
6:10
and then by like january 1st 2020 i said
6:16
i i have like i don't know i had like a month pay left on this grant i'm just gonna quit all my jobs work only on this
6:22
thing it's great um at that point simon jumped in a little bit he described his job as well and said i'll work for a few
6:27
months for free i think it's cool project so samurai jumped in the two of us are coding this alone and basically unfunded in the you know pandemic
6:33
happened and you know no one was around and it was like icf was a meltdown mode the whole time like i was like okay hey
6:39
are they grant turning out people like it the prototypes working people excited about it like you know starting to build a community can you give me like a
6:45
larger grant to build it out more and they're like sorry you're two down the list we have a meltdown jane ethan
6:50
buckman aib is collapsing uh you know we don't care about you basically like come back in six months right which they did
6:57
six months later i came back and they actually gave me some money which is cool us the money to build this thing um for the rest of 2020 but it's kind of
7:03
hard to hear that when you're like dude i'm broke i'm burning my credit cards down and i've done this cool thing everything's just the future of cosmos
7:09
is awesome but like i can't get any cash for it so that was like a mad dash of like
7:15
yeah i don't know i don't know why i didn't give up then honestly i don't know i must have been like just stubborn
7:20
and crazy or something so i guess now the like the the landscape has changed vastly especially
7:26
with um with uh getting grants for for your work now
7:31
right like since those days that you've had uh obviously newer networks like juno and a
7:38
lot of other networks come on that are using cosmosim and sort of depended on its development
7:45
and i imagine they're like you know tipping into the pool to to help that keep moving along as well but they've
7:51
got the um the interchain foundation now too right which which uh helps out in that regard yeah
7:57
well the icf is the irony foundation icf or something wait wait wait sorry i in into interwars i'm sorry not
8:03
entertaining yeah into watson doesn't exist so all i got now we're getting drama that's a drama sign so yeah inner
8:08
chain was really helping out the beginning um then we hit like uh got t grade gave us money and uh then we
8:15
got some from tara even osmosis now juno as various users of this have given us you know significant chunks um
8:23
and i'm definitely very appreciative all these grants so it's kind of interesting going around knocking doors and seeing how to see where the grant
8:28
can come from this six months but there's definitely doors open so it's really friendly i've been feeling really since we've got established this people
8:34
picking up since late last year it's definitely like there's lots of opportunities and lots of people very
8:40
very happy to keep us working um but enjoy some dao doesn't exist interesting it was ideal we launched a
8:46
year ago so there's like this idea of building that out and so we had uh our [ __ ] who's
8:53
worked with us went off to spin it into dao i had people that had like pledged mon vocally pledged this stuff they're
8:58
going to build dow dow dow to do it and i don't know what happened honestly uh the idea is out there
9:04
to fund these grants and stuff like that not for us as much as from theo but for like a lot of other people building cool
9:09
tooling out there a lot i think of what juno ended up doing like juno hat juno and uh and stuff like that so i think i
9:16
don't know he ended up basically working juno projects and then i think just never seen me with this i think he is hacked you know and the junior dev
9:22
grants was enough for him and everyone else so yeah kind of to meme to awesome
9:30
so just a plug right talking about um grants and uh and you know development
9:35
and cool projects right so obviously like you say uh cosmosome was born out of a
9:41
hackathon right so anyone out there listening with a cool idea
9:47
just know that there is good money available in hack juno so
9:53
if you're if you want to develop something awesome and um you know you want to get it funded there's there's various funding
9:59
opportunities in juno and if you've already built something awesome there's hack juno to get a reward for that and
10:06
keep you going so just a just a little plug um
10:11
because you know the whole landscape on funding has changed and there's a lot of it available around the place so
10:17
um definitely know that so sorry i had to throw that in there no it's good it's an important thing i
10:22
mean the whole landscape has totally changed i think this is kind of crazy because it was like back in 2019 when i did the hackathon it
10:28
was the casa's hub it was the council's hub that was it right yeah and there was a bunch of projects they kind
10:36
of waiting for to launch waiting because they come out there and they were i think the few came out
10:41
that year and then they're all sitting there in their own little silos because ibc wasn't working and it was only really
10:47
uh you know in being of 2021 when i received him out the colleges really took life so all for
10:53
a while it was like there's just a hub and then a few other people doing their things everybody took off and then when the ibc came out that
10:58
whole like explosion that was a year ago a year and a half ago the cause actually could scale they actually okay now i'm not just building
11:05
a little machine nowhere i'm building this whole concept so until then it was really only the icf that's fighting anything
11:11
like anything anything and the community cosmos community pool it out like so many things it's amazing to see like how
11:16
many people change their own dev grants their own programs their own community pools um i think i mean juno seems the
11:22
most just focused on causing awesome cool stuff so i just threw out that dow dow
11:28
for me is a really cool project you guys somehow find finance because it's like i wish to build this and never could find
11:33
grants to build government stuff ever it's like something that is needed but no one ever pays for it like every
11:39
patient defines it or like faster databases or like i don't know some some
11:44
ledger signing app but like governance is like always like super essential foundational stuff but
11:49
like always the bottom edge list to pay for so i'm like cool that like the genie community came together and paid for
11:54
that be awesome i love that like to that end that they've more or less taken
12:01
you know the the early success of juno and then just kept driving home more development around you
12:07
know things that they think are gonna um make a difference in the future especially
12:14
um you know with jake and dowd like he's always driving that um so you said like back in the back in
12:20
the start it was you and simon uh grinding away i'm curious was that like did you guys meet up every day and and
12:27
talk about ideas and go through it or was it that like an internet relationship that you guys were grinding through that through github
12:33
uh it was the internet we knew each other like i met him a few times he worked another company for that together remotely we met up like i don't know
12:40
four or five times impersonating me before that and um i think we met in
12:46
in november of 20 2019 actually last time i saw him before the whole pandemic
12:51
craziness started and then it was like it was cool he talked about it he liked the idea of causal laws i'm hacking maybe a few months like i have some free
12:57
time quitting my job soon like let's let's do this together i'd like to work out a little bit and um
13:03
yeah and then they did it and basis all just yeah yeah really i don't know what do
13:09
you have we'd have discord i don't even know how to communicate we'd have slack back in the day maybe it was discord i'm not even sure
13:15
how we dm'd each other it was just maybe the story of github i don't even know if we did any like sort of messaging except
13:20
for github honestly like wondering if people still use the slack is that something that people
13:26
still use yes yeah as i've discovered in the last two weeks yes
13:32
i've never used it in my life uh yeah maybe a orc too that's what mud said yeah it's probably right
13:39
so ethan you say like everything kind of kicked off about um a year year and a half ago right which is when heavy sort
13:46
of development started in the whole cosmos ecosystem i think like probably
13:52
that's a so you might have the same story as this you super that um that's about the time that i
13:58
found the space as well is that around when and maybe you two shields i think the fray might have been around a little
14:04
bit before that or maybe even found at the same time we all in the same boat here we all come in about a year and a half ago
14:10
yeah about feb feb a little bit earlier 2020
14:16
yeah i think i landed in march sorry march 21 is when i landed
14:22
i think that's about whenever i got in as well yeah i yeah i think it was i think it
14:27
was definitely coveted times i remember that is kind of where i started but i don't think i started validating on osmosis until
14:34
21 i think right after right after that went right after that genesis i think that was the first one i was validating
14:40
on what was straight to osmosis really yeah and kicked out
14:45
wow how do you guys think like um
14:50
do you think that we would have had the same sort of uh acceleration in in development and um
14:58
and you know interest if it wasn't for covid do you think that sort of kicked the ball along in terms of crypto and uh
15:05
people's interest having nothing else to do a little bit but i think in cosmos ibc
15:11
is also like not to be underestimated so i i've held co i i held atom
15:19
from like 2019 maybe end of 2019
15:25
and you could see like the interest and the volume like really go up after ibc happened you know
15:33
um but but it did actually like it like a lot of crypto it dipped off at the
15:38
beginning of the pandemic and then the price went up because people were sitting around at home messing around with crypto right that's right
15:46
so yeah 2020 was interesting actually the whole that was like this whole d5 summer wherever in
15:52
this d520 in ethereum that really i think for me it was like we're sitting around and it was kind of the doldrums of i'm tired of uh
15:59
what was it adam's sitting around three four five dollars with adam for a long time up and down
16:04
like three and five and hit six you're like oh yeah it's high oh down again and then once that hit that summer actually
16:10
came out in like eight and stayed over eight i was like excited i was like over eight um
16:16
actually money flow again i think that whole like last part of 2020 was like this side effect of ethereum kickoff
16:22
there's kind of this like all the people coming to crypto ethereum's taking off the pandemic comes at home playing with crypto it's coming coming up and this
16:28
that was wave it went up like eight or ten based on that way i think it was ibc that that really just like threw another
16:34
gear it was good good timing for it good timing five you see i definitely
16:39
yeah because like i was going back through some notebooks recently i was throwing some stuff out and i found a
16:45
journal entry from just before covid um when i saw the last time i saw jake
16:52
before covid and i wrote something like jake won't shut up about cosmos
16:59
but i i bought some atom a while ago maybe that'll shut him up or something like that i had this quite pithy like
17:04
line or in my notebook and so like i have this really distinct like just really like top of my head
17:10
because i saw it a few days ago i was like oh so that's what that's how long i've been holding at him
17:16
to be honest it says jake hartel hartnell got you in there yeah yeah because they were he i think
17:22
it was him and shane were running a validator on the hub right
17:28
uh they're looking at stargaze too they think he started it yeah yeah it was around it was around that sort of time um
17:35
he started saying to anybody who was who knew him oh you should run a validator on on on cosmos hub
17:43
and i looked into it and i was like okay yeah just just like grab a bunch of servers and spend tens of thousands of
17:50
dollars to buy in or something and i was like you know i was i think was that
17:56
i think it was before i started contracting properly i was just like this is a lot of work and a lot of money
18:02
i'm gonna let the people that know what they're doing here continue to do what they're doing
18:07
how much has changed in life i was gonna say didn't you just do the same thing like two weeks ago
18:15
so [Laughter] it's not that is it you can restate the same exact
18:22
thing for another project i think that just started a couple weeks ago yeah but that's just i think that's all of
18:27
technology isn't it like you you you hang around in any space like as
18:32
an engineer or as a no as a node operator in our case at the moment or as an ops engineer or anything like that
18:39
and you just you're on a constant cycle of maybe a year to two years where every two years
18:46
you're back in the same boat feeling like a noob like what the [ __ ] is all this stuff i don't understand anything
18:52
and just needing to confront your own inadequacies again that uh that that is
18:58
that is the the mental health conundrum of working in technology i think uh i
19:04
think this is sort of funny well some engineers front-end engineers have to deal with that back-end
19:09
engineer's not quite as much we have like a like a 10 euro turnover point i think
19:14
i don't know do you want to write me a do you want to write me a startup using ruby on rails
19:20
all right i'll take it all right all right i mean but even even in like what what
19:26
you know because i i've done a lot of fp which i i don't [ __ ] shut up about i know but like you know when i first
19:32
started working in enclosure modern lisp on the jvm 2014 you're like jvm language fine cool
19:40
that's still a thing people do you want to build a closure project now you'd be like it's closure script
19:46
node or nothing you'd just be like jvm are you [ __ ] kidding me i'll just write it in rust
19:52
instead or do you i mean like i feel like even back end languages have changed quite a bit
19:57
in the last few years i think rust is a big driver for that like the ergonomics
20:04
you know it's versus funny thing you love it or hate it like people see it and you smash it against it and you run away once you over the hump it's like
20:10
it's such an amazing language um i was the first dude the first time i tried playing with it i was like this is hard i felt the compiler
20:17
and i sat down actually like i had to like i did tutorials but i hadn't really i couldn't until like it caused some
20:22
awesome like me learning rust as well um so that was a fun experience
20:28
like not easy one but like it was funny after six months it was like i really felt like it became natural like it
20:34
wasn't just i could code it but they came natural and then like the compiler is my friend all the compilers do is
20:40
help me write better stronger faster code just makes everything better it's not hitting me and yelling at me i'm
20:46
telling my homework's done bad it's just helping me it's training me so at that point it was like something
20:51
yeah there's like no turning back like it's like the most amazing language because it's like keeps you doing the right thing it tells
20:58
you it should be doing it helps you so yeah it's like at the end it's super super compact and concise and efficient
21:04
so yeah so um i've got since we're kind of we're heading back in the direction of causing
21:12
wasm now by the sound of things so we're talking about rust um i've got a practical point so ethan you you obviously have done a lot more thinking
21:18
about this uh certainly than i have um i'm kind of curious i i've wondered
21:23
this quite a bit like what is it that drives um because obviously you know look at look
21:28
at aptos um polka dot uh obviously ethereum itself obviously cosmosome all of these are using wasm as
21:36
their target right for the vm what is it about uh wasm as a target and
21:42
like as a runtime that is so that makes it so prevalent is it just
21:48
the availability of it is it just the ergonomics of it because
21:53
again i guess like naively you take a step back into sort of um i guess programming language theory and
22:00
start to think about asts and all that sort of stuff and you go like all right is this the is this the best way of doing things
22:06
rather than because that's that's um jay kuan's big thing with no lang
22:11
right is is is being able to actually represent the code in a in a non obfuscated binary as as i understand it right so
22:18
there are some people who are kind of uh you know whatever whatever you might think of the no lang project that no
22:24
land no lang whatever there's no lang granola yeah yeah
22:30
yeah the gnomes lang um there are but but what i mean is like it it's so much more common to see smart contracts and
22:36
that sort of stuff with wasm as the target like for somebody who's not
22:41
written cosmos and why do you why is it that wasm is always the the kind of target for the smart
22:47
contract well i mean there's two ways of doing it there's basically you have the choice of doing a
22:52
vm where you compile any language to the vm representation uh vm is basically a virtual cpu a
22:59
virtual machine right so it's some target you can pile something down to and
23:04
like evm or the jvm your favorite one over there and um the other pro interpreting it so what granola does
23:10
interprets it's actually pretty interesting um i looked at a project there's some really cool ideas he has about like how map storage and actually
23:16
starting whole pointer sequences and all the mercalizing memory um it's doing some stuff kind of like the
23:23
active records from ruby and stuff like that because it's all interpreted language um which is could do more powerful stuff but for me
23:31
you're kind of embedding the whole interpreter run time and your programming language into one one blob
23:37
it's kind of complex and then you have to write the whole thing maintain the whole thing right um taser i see that one as well
23:46
they have a language you compile you can run the tasteless language right i think they don't vm for it just have a language interpret i think a lot of
23:52
these ones write their languages of grow language that was coming out there all these other ones invent their own
23:57
language and interpret for it this was like the first thing everyone did um except for ethereum um was the one a vm
24:04
and when a theme came out they looked at basically we want a solid virtual machine to run on right and they look that they have all their hashing things
24:10
they want to do the hashing stuff for their the rules and the time if you wanted to use a drop
24:15
in vm to use it was a jvm if you look at that thing comes this huge history not just like but like it's a whole run time
24:22
like the georgia virtual machine is tied to java language and or like well derivatives kotlin is nicer actually or
24:29
you know closure or something but like it's also tied to entire system calls and everything like that you can't just
24:34
embed it somewhere else um the evm says hey we have to build own virtual machine
24:39
and around that time this was coming out of uh mozilla and stuff again web assembly
24:46
and it came into a virtual machine the virtual machines were simple rather make it more complex with the opposite approach of jvm which was like let's
24:52
make a simple one it's just a simple cpu an old-school cpu and memory nothing else you can't read files you can't do
24:58
the network you can't print your screen you can't have a gpu you can't do anything really right
25:05
so you can't even sleep five seconds you can't launch a process you can't do anything all you can do is actually take
25:11
the input and run a function and come out and the inputs can only be integers right you can't even take strings and take integers you map everything else
25:18
so it's really really simple and what that meant is you could add but then you could add customizations to it so you could customize it right so it's
25:24
basically a super simple thing train down and customize too and so anything you want to do that you can
25:30
extend it to add your custom blockchain stuff to custom mods extension points here entry points here
25:35
custom calls to send messages to query people right to check my values
25:41
but everything else is done for you so if you think about it a lot of different teams work to make
25:48
really efficient watson virtual machines not just the the browsers but also
25:53
awesome time and wazimer and a bunch and and go wagon a bunch of other ones implementations of this very simple spec
25:59
through a faster and faster ones you have a whole a series of competition to build the best possible virtual machine the
26:05
fastest one most secure one right and then you have many different languages that exist in our really supported
26:11
mainstream languages to compile down to this virtual machine so c and c plus plus the first ones russ came up very
26:16
shortly as amazing support there's also some support for other things like i know goes working time ago
26:22
works a bit on that one um the bunch of esoteric languages also going to this thing like v-lang compiles down to it
26:27
and nym i think um so there's a few you have a choice basically you can take a handful of existing full-fledged
26:35
languages that other people build you don't build your own language like a celebrity you have another existing ecosystem language you have a virtual
26:41
machine that a lot of your teams are maintaining versions of make them faster and faster right and then you just function on how do you
26:48
plug this virtual machine into a blockchain that saves all that [ __ ] work for you
26:53
don't go the language the tooling around a language id plug into the language uh
26:58
try to write your own virtual machine certification optimize virtual machine specification work in edge cases on it
27:04
run git compilers just time compilers et cetera et cetera and like at this point
27:09
water is fast and you don't even worry about not just literally like trying to figure out how to do the gas cost is because like it just runs so fast
27:14
because you have like the speed of rust the street of highly optimized jit compiler on the virtual machine so
27:20
basically you get for free so everyone says hey we got all this amazing super free let's just plug into a blockchain
27:26
and worry about like apis are exposed on the system rather than building this hard system you know
27:32
we're not computer science because i guess even if you had uh from a plt perspective like
27:39
another um if everybody okay so basically like if everybody wasn't so allergic to
27:45
parentheses the the obvious solution would be a lisp because every single lisp is self-embedded right it's it you
27:52
know that we've all seen sicp where he comes up with the with the fez on his head and
27:57
says i'm going to show you the i'm going to show you where the magic happens you know like the the kind of meta circular compiler nature of all
28:04
kind of list programs where the ast blah blah blah my economy stuff but
28:09
the thing with that is that even if you did that it's not going to be an optimized run time is it right whereas wasm like you say
28:17
the performance has been squeezed out at this point like it's used by a lot of people for for
28:23
performance programs and there's additional optimizations you can do in the language that you're you know on its
28:28
way down to that compile target right
28:34
yeah i mean i don't know lisp i think everything is a list
28:40
i learned that in school and then kind of ran it killed function programming for me it took about 10 years and 50 years old functional programming after
28:46
that i think when you said earlier you were like oh people either love or hate rust i was like no no people love or hate
28:52
lisp that's the one anybody can learn rust you just bash your head against it a little bit and
28:57
the types will make sense it's not haskell it's not that bad but lisp is one of those you either
29:02
experience it and you're like this is what i've been missing in programming or you're like i [ __ ] hate it um yeah
29:09
and yeah it's i don't know it's an addictive drug if you're into it um i tried haskell and it made sense as
29:15
long as doing math as soon as i said okay i want to read input for the user i want to write data the database i want
29:21
to uh park the network oh that's complex esoteric stuff you can't do that just do
29:27
math no input just oh but just like you're very pro given this math for you it's
29:32
great but but at the same time as soon as you get past the as soon as you get past you know the
29:37
people [ __ ] posting the oh what is a monad it's just a what is it a monoid in the category of
29:44
ender functors what's the problem you know kind of [ __ ] posting and all of that you go and the burrito thing as
29:50
well which is the single most unhelpful thing that another programmer has ever said to me
29:55
oh what's the problem is just like a burrito yeah okay thanks man that's not going to [ __ ] help me implement this thing this week um
30:01
but the idea that you have composable boxes of logic that are mathematically provable i don't have to have a maths
30:08
degree to believe that the monad proof holds right so that that is really powerful but like you say
30:15
do i should i have to worry about that to do i o probably not but then the flip side of
30:21
that is in rust you do have to use a monad to do i o right you using a result type or a
30:28
wrapped result type but nobody knows it's a monad because that's not how rust puts it to
30:34
you so it's again it's like it's that ergonomic that extra bit they've thought about the
30:39
language they've gone like should we just give everybody a result type
30:44
and not freak people out or shall we tell them it's a monad it's like no no just give them an option a result and
30:50
let's just move on with our lives and let's give them an npm style build system that's nice and modern and not
30:55
[ __ ] like the high school build system um yeah like the the the the first time i
31:01
ever used a haskell compiler and it was like this isn't going to compile mate but i think the code you meant to write was
31:08
this i can just like fix that for you because this is long before github
31:13
co-pilot i was like holy [ __ ] and then when i saw that with rust and rust and clippy and that sort of stuff i was like
31:19
yes this is great sick i'm an idiot i like i quite like the
31:25
computer telling me you're wrong here's why you're wrong i've done some working out i think
31:30
here's the solution it's wonderful which we says you're wrong with another
31:36
message that sucks error but like what says you're wrong this is why you're wrong and here's what
31:41
you should do and here's a link to an article explains it if you understand it more like i'm like wow thank you master
31:48
exactly teach me teach me students
31:55
before you got into before you sort of got into crypto ethan what were you i kind of i think i remember the answer to
32:00
this but what what kind of software were you working on building so like what's your what's your kind of background in
32:06
work in that is it kind of low level stuff like uh like drivers and stuff or is it
32:11
something else yeah there's web stuff a bunch of products so i mean um and all kinds of stuff through the
32:18
knocks i'd say a bunch of random jobs the i worked at ableton live i don't know if you ever worked with that dj
32:24
stuff cfc i [ __ ] love ableton oh my god it's it was like the only decent
32:30
technology at the time it was an awesome place i [ __ ] do you know what actually the other thing that i got the the time
32:38
where i wrote in my notebook all jake will do is talk about cosmos the other thing that i got him to do
32:44
that time i saw him was download a copy of ableton live because he was like yeah he was complaining about the ergonomics
32:50
of logic i think and was it was doing some demos for some stuff and i was just like just [ __ ]
32:56
download ableton dude just download ableton it would change your life he's a [ __ ] starter drug to buy all
33:02
kinds of midi controllers like that's really what it is i needed the keyboard no i need a pad a drum pad i have a
33:07
little like the boxes there's so many different boxes that they kept giving us like samples of stuff there oh my god it
33:13
was cool it was cool good place um
33:21
okay so it's music related gotcha it's like i was listening i was working at ableton but then i was also end up
33:28
doing other more like um back-end it's back-end for uh web stuff there's a bunch of back-up web
33:33
stuff after that the biggest order bird and yeah basically an iphone app that
33:38
ordered for restaurants ordering stuff whatever we had big databases thousands of restaurant data in there and stuff
33:43
like that and api services doing all the back in api isn't it um and there was really scaling it's
33:49
actually funny before i got into blockchain it's like trying to figure out how to handle conflicts between like multiple people
33:54
editing the same day at the same time and you want to not lock it without locking so you got those weird concepts of that like crt before the blockchain
34:01
so it's like getting nerding out on like message queues and crts and
34:09
but how do you allow two people to edit the same thing at the same time without crashing and breaking so like yeah um
34:17
anyway it was it was interesting an interesting thing and i was kind of in that rabbit hole anyway of like trying to figure how much scalable back-end
34:22
systems in this case is before i got in there so like i remember looking somehow at a
34:31
raph consensus algorithm for some other reason i have no idea why before i even knew about tinderman and i found
34:37
tinderman i'm like okay this is like you know okay this is a much more advanced version but like yeah so i was kind of getting i
34:42
got somehow into that weird weird loophole of backhand engineer getting too much into their into the theory
34:50
yeah i guess that makes what i mean were you were you guys running like kafka or cassandra or things like that to to
34:56
manage all that data because that's i think that's quite a common way into distributed systems for back-end engineers is having to [ __ ]
35:03
around with zookeeper to keep the kafka clusters organized that's certainly what it was for me like
35:09
schultz's nodding i feel like that might be the same uh oh yeah i worked with that very heavily
35:14
when i was at acorns yeah this is 10 years ago we actually used kafka it's kind of like trying to
35:20
uh to get more stuff in there and use yeah even looking various systems back there we were
35:26
talking about giant [ __ ] clustered postgres mango
35:42
not sure it's possible to start it and they kept pushing it back and they kept growing and we're like dude we're like 450 gigabytes we got a shard now
35:50
like they're not gonna lie ever again so anyway that was a uh that was interesting one so yeah lots of
35:56
i would say lots of tech debt in that company yeah i think everybody has a story where
36:02
they're like uh you should never do this but equally
36:07
that is what i that day that's how that's what i went to work and was paid
36:12
to do and i'm not proud of it but we got to the end of the day uh
36:18
i think the work the worst thing i could ever do was we had this we had this massive spark cluster but it
36:24
was paid for like microsoft were trying to get us to use a managed spark service and we had
36:30
unlimited disk space but quite bound cpu space
36:35
you can see where this is going got you yeah it turned out that because it was backed by like azure blob or whatever
36:40
you could just make a massive fake swap uh disk over like all of the azure
36:48
storage that we had in our account which was [ __ ] lots and then just run r basically just run
36:55
all the different things and just have them dip into the the the swap and when they ran out of space when they ran out
37:01
of um a ram or compute and uh that's how we managed to get a
37:07
multiple of what we were supposed to be using on the slightly above free tier job done like a multi multi kind of
37:14
terabyte join basically um across some databases and they were just like oh we're seeing
37:20
a lot of usage from your desks today you're shuffling around lots of data and you're like yep
37:25
yep yep well i got a little note before we started this chat i'm happy to just ge here geeking out but i got some note that we
37:32
should not talk about dev stuff drama per second and some other top points you should talk about i think there
37:38
well i i did actually want to talk about some more wasm stuff
37:44
so look i've got a burning question for myself right and uh i'd be remiss if i didn't argue it's always been uh on my
37:51
mind um a potential uh problem about blockchain performance with validator hardware is
37:58
that like something that as uh the well as was and becomes more and
38:03
more used um actually processing all of
38:09
those that contract um usage inside vms in between the blocks like
38:16
how does that relationship work and like as there's more load on the blockchain how does that affect validated
38:22
performance and and like minimum hardware specs and that type of stuff so are we going to get like dragged down by
38:28
the lowest common denominator on on the blockchain or is it not really of concern it just won't get there and miss
38:35
blocks or yeah how does it work i mean nobody is dragged down if like 70
38:40
of people are fast enough the other low ones miss blocks they get screwed so like you kind of dragged in the masses but
38:46
like five five bad validators of crappy hardware just kicked off basically they'll be jailed because you can't run
38:52
the blocks anymore happiness mostly i think too soon like if you're running in four gigs of ram you'll run fine for a while and then
38:58
just crash sometimes actually comes in and you can't handle it so the crappy ones just fall off basically and all the
39:03
mid level basically you settle around the mid level right so as long as 70 you know the bottom 20
39:09
percent are you can you drop off but if you're down to like you know the bottom thirty percent maybe forty percent like
39:15
that three percentile is what keeps it uh locked down to that one um the and i looked at this i timed this
39:21
stuff we did a lot a lot of work and performance and so wasm is not that much slower than native code running cpu time
39:29
talking about uses about memory use a lot more memory actually uh but not much for cpu time but it's
39:34
sort of storage and all the read writes of storage for disk and so um i looked at stuff and basically
39:40
that's just if you want to go faster than blockchains just speed the storage just just that's a that's not just but
39:46
like yeah just storage access um you have good disks you want good ssds but like speed up storage reading and
39:52
writing that will go faster and that's and that's that's i have all right because that's also it's got its
39:59
own its own um yeah yeah i wanted to refactor that 2017 as february 2017 i
40:05
was trying to factor and realize how slow it was we should rewrite this thing not rewrite it it wasn't bad code it's just like you can clean it up and like
40:11
optimize and all these things like i pull up that project so now it's now it's actually priority it's fine it's too early then all these things i want
40:18
to do by 2017 is still too early and they're actually probably due 2019 and they're kind of getting around to doing them these days uh they're trying to
40:24
build a new one smt i don't know who will do what they're supposed to try to tune up ivl their alternatives out there
40:29
i don't really know what's going on no one's really working on it that much i think so um it's i guess it is pretty it it's off
40:36
the end of quite of quite hard and into very hard some of that stuff yeah you have to really low-level people
40:42
i mean yeah and so they're all doing other stuff so the problem here is basically we have infrastructure of his product
40:48
and like while everyone's sitting around there like and adam was like you know nothing's really going on until ibc came
40:53
on it's like there's all this money infrastructure and i like oh with money money money we're gonna go slow pace no rush and
41:00
then the money comes in and like last year it's like okay money build build build no time infrastructure every dev that's skilled is building right
41:06
everything has the skills is building product product product product product and there's no time for infrastructure now everything the university should put
41:13
last two years we need it now so we don't have no one's building infrastructure and we can repair the limits of the
41:19
instruction you have right because it was actually more urgent than thought so um it's a weird situation now we're in
41:25
the bear market so hopefully some of these devs build the infrastructure we need and get ready so you know six months a year from now we'll actually
41:30
have see how the infection we need um so that's so that's a little bit on
41:35
to dev on boarding and that sort of stuff as well which uh we've got another question on here which i'm going to
41:41
segue into which is sorry not i'm still not done okay
41:49
validator onboarding in a minute so related to the performance um ethan i
41:54
was wondering uh so this in certain um environments now
41:59
uh for certain applications like there's a bit of a race for faster blocks um so you know we're getting down to
42:06
like the half second block range in some networks for networks that use cosmosm
42:12
um is like how will that affect the uh like lowering we're speeding up
42:20
block speeds um so there's like more more blocks per minute is there going to be a point
42:25
where there's just not enough time um to process contracts in between actually
42:31
minting blocks and if if so does that just mean that all of the the cosmos and transactions just don't
42:38
make it into blocks so how no it just slows down you have to make a block so you have to make a block and then you
42:44
have to run the block then you can't make another block you run that block attendant works that way turn it works that needs to finish if you make a block
42:51
x they never have to run that one and the result of running it is actually part of making block x plus one
42:57
okay so you you have to finish the uh you have to finish writing it before you can start x block
43:03
um right i've turned down i'm playing around local local nodes you tuned it down and i'm running twenty uh one
43:09
quarter second blocks in local no one node is running contracts they're running i turn the consensus down to
43:14
like i think 200 millisecond pauses and it's running every 250 300
43:20
milliseconds is popping a block out there with contracts ibc transactions runs um i think some of you have they have a
43:26
bunch of small blocks it's fine rather than one big block right so the execution time's not any different the
43:32
main time is actually running this block is a lot of pdp stuff i i think that what null is getting at
43:38
is that that's by turning down the iota um not the iota oh there's one thing
43:45
it's it's yeah it's it's by now you can commit timeouts and right it's by it's by tuning that
43:51
configuration to to bring the block time down but i think a lot of chains are bringing the timeout time down
43:57
on a per validator basis to speed up the blocks and i think what null is getting at is that if you
44:03
are bringing down block time using the timeout then isn't there a chance that you'll
44:09
end up with higher gas or slower execution like larger things that potentially just
44:15
not going to make it in a block as regularly if you see what i mean yeah that's i think it's a bad decision there
44:20
honestly i think there's just one and a bunch of designs honestly like i i think the proper design and i was
44:27
pushing this i built this back in 2018 it's another my other sdk uh so i built like between i built the
44:33
cosmos aka by the way in 2017 like zero six zero eight if you look at like if you look at main cause it's game one of
44:39
the big contributors was really like that summer of 2017 um and they took off they built on it um
44:45
and then i got one off my an sd case i thought was better um and that had
44:50
things like um like protobuf types back in 2018 and a javascript client they could actually
44:56
sign messages without using some other go medium stuff and multisig kind of dial basically there's
45:02
getting of dow dow was kind of in there at least uh you know group contracts um or all these features there was nice one
45:09
thing we did there is called weave um and one thing we did in there was also
45:14
work on this uh no empty blocks it's a really cool thing which is adaptive block timing so you
45:21
can tune your system so uh it will have a minimum block time of half a second
45:28
but if no transactions come in it skips it so if so
45:33
the blocks are lazy basically it tries to make you tune it it should work like every every second or let's
45:39
say every every one second you try to make a block but if there's no no changes in that block then just skip it
45:44
and wait and then say well at least every five minutes you make a block right so at least one every five minutes comes out there um but if there's if
45:52
there are transactions she'll make every second comes a block and there's a pause and maybe seven seconds and then another few seconds a few blocks come out and
45:58
then pause i was talking to someone about this the other day the that same concept of like
46:05
variable block times and the the issue i had in my head is actually
46:11
for inflationary chains is the the minting because you would have to then vary
46:17
it like it would be really yeah this doesn't mean
46:22
the whole the basically they built in tenement a while ago and entire sdk development as kill this whole idea
46:28
first of all um the way they store things they always write something every block there's nobody blocks ever
46:34
effectively they change state every block so you need blocks they don't change state to make it work because they write stuff every blockchain pos
46:40
stores lots of rights every block um the second thing is shooting epic like expect payments if
46:46
you have epic rewards being dumped then you can do it the other point is they
46:51
measure everything a number of blocks not in time that's historical because back in the day
46:57
like in 2019 there's no guarantees or 18 on tendermint that your block time is actually really trustful and you can
47:04
play with it so um in [Music] they got that in there like in 2019 or
47:10
2020 that the the block time is guaranteed they see byzantine fault block times
47:15
so at this point honestly there's no reason to give a reward per height per block we should do it per hour or per
47:22
day or something um if you look at t grade they do that we just do it for time we do it for time not per block
47:28
so there are like a lot of complexities in using time-based things in
47:35
in blockchain though right you can't use the real time but the block time so block time is trustable
47:42
just like height is stressful it's always going up there's a lot of rules on what block time can do so you can't use the current time what time is it now
47:49
because my computer your computer are different and not just different
47:54
time zones but actually you know might have off by a few seconds or milliseconds but someone might run that
47:59
same blockchain five years later and the time then is five years different than it is now by the time the block was minted when
48:04
you create a block you put a time on that block which must increase every time and two-thirds of the values must agree
48:11
that's around time within like a few seconds so you can shift it a little bit if you're playing per second you can
48:16
play with it a little bit validators can shift things around and play a few seconds um
48:23
but you can't really push it much you can't like shift your hour ahead something like that doesn't matter what you try
48:30
unless you own unless you fork your chambers that's enough to fork the chain you can't do that yeah so so minting
48:35
would have to completely change for it to work otherwise people would just
48:40
like if if you're if your actual if you kept going per block for your
48:46
emission and you wanted to like say like you were saying have if there's no
48:51
transactions then only do a block like every five minutes or something then people would just game it and smash it
48:57
with [ __ ] transactions just to keep them they keep getting their rewards coming so
49:02
that's awesome but it's all it's all doable i'm not saying it's all this is doable right like so um
49:08
it's it's been done the technology is there it's design of the pos system and so all the shield few things here right now
49:15
so i think the pos system was interesting thing that was done in the hack and if you you hear about the whole collapse of aib
49:21
end of 2019 2020 or did you miss that drama for a second
49:26
so they basically say it was a chaos death march polarization uh
49:32
and some stuff was going on basically there's a bunch of people hacking away and they got something working and pos in cosmos is stable
49:40
implementation is stable not designed that's quite nice the implantation is stable
49:45
i think largely because it's been tested so heavily and it's been like fuzzed in production you know so it's stable
49:52
um but i don't think it is a best implantation possible and it's so hard to change it's like i've looked at
49:58
code it's like it's scary because you can't really understand it you'll have to change it but it works so the liquid sticky module has taken them like a year
50:04
and a half with a lot of really top devs stability because they can't touch this module right so i think it's time to actually have a different module so like
50:10
t greater perfect always build their own own module of our own sticky module and distribution module and governance
50:15
modules using actually rush customers contracts for everything this is all built in cost multiple contracts
50:21
um so pretty powerful ethan we had a question actually so i'll just pick you up on there from rama who
50:27
said when will we get a wasm sdk or was some smart contracts as modules
50:34
so i was going to say to rama let's get ethan to talk about t-grade so now you're talking about it um can you
50:40
explain a little bit about how that works um for those that maybe haven't been following t-grade or t-wasn't
50:46
cool yeah t-grade basically has um having a product here so uh t-grade
50:51
is basically we add hooks so modules can call in to privileged things you have
50:58
privileged modules then you click in begin block and block control validators adjust consensus parameters there's
51:03
various things that modules can do that contracts can't do you can promote contracts to be privileged or right
51:09
because the modules so after that point you have to set up mod go code you set these up modules up
51:15
lie them out there then you have a validator module just sets a value set every block and you or or not
51:22
you have a staking module um and you have it can begin block things so it can
51:28
automatically pay you back by the how do you claim it it just checks every begin block who's out to pay out and pay out
51:33
people their rewards um and so they get hugging to begin block and block they have a governance voting
51:39
modules so the values can vote and actually set up an x upgrade call it's called x upgrade to set the chain upgrade
51:45
they can change 10 parameters so you can do all of it we just wrote that as extra bindings custom message bindings
51:51
basically it works the rest of the way inside of your own custom message bindings otherwise the normal class of awesome
51:56
contracts so i think it's really powerful and one thing i also mentioned there's a bug
52:02
right now actually it turned out that someone tried to withdraw i guess zero i mean check the zero and then they try to
52:08
send a zero claim in the and begin block and that caused some sort of you know error which we just bought a bank so
52:15
that blocked everything else but now no one can claim i think it's still out live now it happened a few days ago um
52:21
like but the same day within finding it within that was yesterday actually within like an hour uh we identified it
52:30
with another two hours on had written a new code to do it um
52:35
to fix it on a contract and then you compile a new contract now you do upload the contract i just have to vote so at
52:42
that point you have to start the blockchain or anything you can write the contract upload a contract and then the values can just have a normal vote like
52:49
a dow down vote right to swap these swap the modules hey do a migration from the
52:54
old one to the new one so you don't actually have to reset the chain for almost anything you can adjust the core levels of consensus
53:00
just by doing a contract migration i mean quote just because you actually need to have the values that has to vote
53:06
on it the gov contract for it right so it's a very important vote um but yeah you can you can do that stuff you have
53:12
that right privileges and it's all then i think and the governance we've seen doubt our governance versus native
53:17
government is pretty powerful so i'd love to actually put some more downloads up there we've owned like it's not down out of our own like kind of similar
53:23
ecosystem there that is really cool that's that's pretty killer actually
53:30
that that particular thing um the fact that you can upgrade modules without
53:35
having to you know have a have an upgrade with everybody and everybody has to do the upgrade
53:41
is like just through change swapping out contracts is that's pretty killer i have to say
53:49
that's where 90 of the [ __ ] ups happen during upgrades right
53:54
like case in point the other day old mate has double signed during a [ __ ] up upgrade oh actually no it's
54:00
just sorry that's wrong it was just during uh business as usual right and uh there was a problem
54:07
which resulted in us a halt which needed to get restarted and
54:13
but they did have to upgrade the binaries which then resulted in some database issues which then resulted
54:19
in loss of state i'm assuming from the the signing state which resulted in a double sign so
54:26
um once again a non-malicious double sign but a double sign all the same uh so i think they've had an upgrade this
54:33
is a secret by the way they've had an upgrade to like um well they're voting on an upgrade now to restore
54:39
the validator and the and the slash but um just interesting that that that
54:44
type of system can then be in place to just for most upgrades just walk right past
54:50
all that stuff and then you know you vote on it it's done
54:55
so i'm aware we've got about five minutes left with our allotted hour with ethan so uh as a kind of
55:02
pushing pushing things along as a kind of towards a wrap up maybe so
55:08
before we push things along to the wrap-up um the fray can you just so everybody knows in the audience can you
55:13
actually tell them the the format uh well okay yes it's somewhat someone
55:19
takes that so we're experimenting with a new format where we're going to keep it to an hour with our guests we're not doing any news or anything like that at
55:25
the top of the show and we're going to have less structured stuff after we've talked to guests because we need to be
55:32
more polite with our guests time and stay on topic which is hopefully what me and all have been doing
55:38
with the questions we've had during this talk and by only interrupting each other and not our guests um but we might we
55:44
might want to just give maybe an extra five minutes for ethan to actually talk about wind because it's pretty important for him so
55:50
that is also true um but so on on just on the cosmos more
55:55
generally ethan like uh well and i suppose mainly cosmism really in this context um you know what do you what do
56:02
you think the biggest challenges like facing the cosmos are um so like
56:08
going forward like where do you see the ecosystem going and what what is causing wasm's role within that
56:16
so i think cosmic cosmos causing water from cosmic coming together more and more right like it was a few change and
56:23
now like i think most of the major changes now have caused a variety of it the hub and efforts
56:28
don't but i think if you look at the next five or six different important chains i mean you
56:34
look as most of those you know injective secrets um they're all in it persistence rolls
56:40
out i don't keep seeing more portable rolling out custom awesome on there so i think there are non-customizing chains out there i think a large number of the
56:47
important chains are costum awesome so like it's a subset but i think it's important subset right and they're both moving
56:53
together so um i see cosmos innovation a lot of ways you see a lot of innovation happening especially ibc later
57:00
faster than other chains other non-cosmos and chains but i think a lot of the non-plasmas and chains can really go for stability things things are well
57:07
defined problems right if you know a problem is well defined it'll be the same problem a year it's a great thing for go modules if it's changing every
57:13
three months it's a better thing for classroom awesome just move faster so i think that's a big difference but where you see is going to ibc like ibc
57:21
has proven that one year of like token transfers and the whole tokens and connecting bridges and obviously bridges
57:27
holding up while these bridges keep falling over like i think is proving cosmos idea but ibc is it's a baby dude
57:34
industry accounts are trying to come out there initiate security interesting queries um those are the beginning
57:39
generic protocols we have hundreds of protocols that go off and riff on those on other you know if you want to talk to me on this we need a whole podcast just
57:46
in this one but like i think um ibc it features ibc
57:51
and multi-chain and going beyond what everything is supporting from from ethereum like i'll play the same app and through change but actually having like
57:57
parts of your have one application architected uses parts of different chains right like maybe it has part of its uh you
58:04
know it's handling like a liquidity pool that's like uh on osmosis which is um using the auto
58:12
liquidating on the decks right and you have another piece it's like um liquidy uh
58:19
sticking dirty launched on juno another piece is like secret voting launch on secret network right you have arctic
58:25
this way and you can you can build something like that and okay now to connect these pieces together and actually don't just have people many of
58:31
them but this this app on one chain athletic chain or talk to each other a special way to think actually arctic is
58:36
out is is very like the next step um and along with that that's just technically i think really what needs to happen in
58:43
cosmos is is users um the devs are there a lot of devs are
58:48
here but like i think it's the unification i see i may start to ibc because ibc means
58:53
people coming together it means not like silo one side of two instead of three fighting but like silent one style two
58:59
three coming together in one larger group i mean sharing users sharing devs sharing mindshare right um and that's
59:06
how like you know no zone is compete with ethereum but 40 zones working together collaborating
59:12
you know it's like the [ __ ] voltron right so i think that's where the future of uh futureness is so you know getting
59:18
more users and really collaborating and breaking these like different silo chains not just move my money into my
59:23
chain on your chain but like really collaborate and have these applications are really truly crosstrained um and with that pull in a huge new
59:31
group of users it makes sense um something something i think i've i've
59:38
i think i've obviously talked on the podcast about before but also i think i think i've talked to you about ethan as
59:43
well is that i'm personally pretty excited about um penumbra and i think like not only because of their model not
59:49
only because of all the stuff they're doing with zk and everything but also because they're an ibc zone that's
59:54
not using the sdk so it's like it's beginning to fulfill the app chain thesis outside of people building with a
1:00:01
monolithic stack which i think is really interesting as well like again for bringing in other cohorts of users from
1:00:07
radically different use cases totally totally i think it's really cool
1:00:12
to build it i'd love to see it actually they push things in different ways it's awesome to see that um
1:00:18
then push things it forces the stack to actually become multi-chain
1:00:24
multiple customers okay but like really multi-chain yeah well i mean obviously i'd love to see well i i don't yeah i don't think i
1:00:30
would love to see a rust sdk but i would like to see libraries available for because there's
1:00:35
obviously the ibc rust library it would be great to i've already bugged ethan about this like how what would
1:00:42
need to happen to have a an ib a cosmos and rust um i'd love to see that but um
1:00:49
ethan windau would you like to just talk about window for a little bit
1:00:54
before i loved it usurper rugs you off of the show i'd love to before i get pulled off the stage i love it rip on
1:01:02
window so uh yeah who knows maybe why i'm not building russia's decay it's like i've infrastructure for five and a
1:01:08
half years and i want to build something meaningful like a product that's not meaningful um so i was watching d5 study all the d5
1:01:15
stuff didn't jump in the fray back then bull market didn't want something and win for me is something close to my
1:01:21
heart it's like i try to build a meaningful product that actually fits the markets and it falls into this like multi-chain
1:01:27
application thesis i was telling you about like actually connecting all the chains and building something that's like bigger than some of the parts that
1:01:33
kind of feeling and technical and also trying to build bring community together cross chain so
1:01:38
window first of all window is right now we have a token on juno and the dao is coming live very soon it's like a very
1:01:45
customized dowdo application it's gonna be live just to know what it is and that dao which if you a juno or osmo
1:01:52
staker you hopefully got the airdrop because in regent stickers we dropped a whole bunch of tokens another round
1:01:57
coming soon but um but it's coming out basically it's on there we are now launching uh protocols d5
1:02:04
protocols cross chain and multiple chains the idea and all controlled by one governance token not like everyone makes its own
1:02:11
governance token a legal structure just have one one thing come in this join the window and we launch lots of d5 protocols so one coming soon is taking
1:02:18
derivatives on juno that should be here within two months so we want to have the native and not have to like send your tokens another chain
1:02:25
just need it on juno we can replicate other ones there's some uh some nice bonuses on this thing but you
1:02:31
know whatever another state derivative we want to add other things like stable coins lending protocols et cetera et cetera oracle services
1:02:38
um we're working all these pieces in design stage the idea is to run out a whole bunch of ones on different chains
1:02:44
and then figure out how to how to like connect them crosstrain to be able to define ecosystem and collaborate with
1:02:49
other projects and once you get other projects out there once you have one dow it's really good the dow about the dow
1:02:55
not about your protocol you can have a dow another down merge swap what does it mean to have like you know
1:03:01
protocols collecting that's moving tokens actually interacting in deeper ways so i really want to experiment with that like experiment with having one dow
1:03:07
control projects and multiple chains directly um and see it bring a huge
1:03:12
revenue into that and also figure out how you can merge dowels how you can do like yours token swap swing dials uh
1:03:18
android partnerships in dallas uh because that's a doubt that's also downhouse model isn't it uh you know
1:03:24
being on multiple chains doing token swaps where necessary that sort of stuff so again like you say these these
1:03:29
applications that are coming along that don't have one just one home chain i'm super stuck in doubt out like doubt
1:03:36
i was kind of like things i wish i had been had time to build and then i found the guys are building this stuff like it's awesome it's like the very same
1:03:42
philosophy i've had about governance and like you guys build this is awesome thank you i like i built i built cw3 and
1:03:48
cb4 back in like late 2020 hoping someone would take it like a year later i found like this down the ui like oh my
1:03:54
god they're actually doing it so that's super awesome um so yeah most would not relate to trees wind is also trees so
1:04:01
beyond building this technical d5 protocol what we're doing is so i didn't
1:04:06
want to build like the cool rbc device stuff i want to do something with d5 like i came in to put out money out of
1:04:12
something better money more meaningful money that represents values rather than just you know just money to
1:04:18
buy something so what i was trying to do that lots of ways what we're doing in the wind
1:04:24
is taking 25 of all revenue from the protocol and donated to environment let me especially
1:04:32
and that would be like a committee appointed to pick projects and dedicate it to things and whether it's clean ocean plastics or planting trees or
1:04:38
investing in future tech trying to green deserts and uh do crazy you know moonshot environmental stuff everything
1:04:45
from environmental good basically is all um oh man everything from there is all
1:04:51
going to be funded by this so whoever the dow wants to vote on is open so at the same time it'll be extending d5
1:04:57
protocols on multiple chains it's pushing cosmos obviously the limits and you know being a big dow-dow is based on
1:05:02
doubt um it's also take the profits of that a lot that so you can basically help the planet and make money at the
1:05:08
same time
1:05:21
okay who did the rugging there but that was very efficient
1:05:28
um for those of you listening at home uh that was ethan uh father of course and wasn't ethan frey
1:05:35
the other phrase the og fray hopefully that was an interesting talk i think i learnt
1:05:41
a lot my head hurts a little bit um maybe we should invite ethan back uh some more questions well
1:05:48
we got like three questions we had like six questions on the list just as like here's some question obvious questions
1:05:54
to ask and we got like halfway down then so right uh and yeah i even got to get
1:06:00
out i even had an opportunity to get out a keytar at one point so if that's not a good conversation i
1:06:05
don't know what is uh again if you're listening back on a podcast player you might have missed that bit
1:06:11
there were guitars earlier so um yeah um
1:06:17
the what would you want to what do you want to cover off first you want to do follow-up so let's let's let's kick off
1:06:23
with our follow-up questions housekeeping and that usually just leads into the [ __ ] anyway right
1:06:29
sure true out you gotta talk about your uh you have to talk about your your gaming nodes uh twitter
1:06:36
uh oh yeah okay so ethan was kind of like uh referring
1:06:43
to uh to this little bit earlier so we put on a poll which about seven or eight people answered no it had it had some
1:06:49
decent amount of answering so ethan was talking about um you know people the people want to see the dps so he was
1:06:55
trying to segue into some dps but ethan that's that's not the main show mate the dps is after probably i mean he's
1:07:02
wait he might want to come back for the dps but the problem is if we he's still here but
1:07:07
the problem is if we bring him up you [ __ ] are just going to ask him about it's you know dev [ __ ] so
1:07:15
i don't know the format's the format now stop complaining stop stop showing how the
1:07:20
stop showing how the sausage is made you just oh my goodness so anyway
1:07:25
we did a uh we did a poll on the twitter and for [ __ ] sake like
1:07:31
just follow us if you don't because we're like six months into this thing and although
1:07:37
we don't market we shouldn't have to all right it's a [ __ ] awesome show and we we only just cracked a thousand
1:07:44
followers i put out a tweet that we're almost at a thousand followers
1:07:49
and i [ __ ] someone left we were at 9.99 and then someone rolled back to 997.
1:07:56
[ __ ] you so obviously or australian cult movies
1:08:02
with me i i was the other day i was like looking for some more people to follow right
1:08:07
because i like to like you know follow people in the eco but the problem is that
1:08:13
i don't want the following to change from 69 so every time i go and follow someone i have to like go through the
1:08:18
list and unfollow someone so if you've been unfollowed from game of nodes lately you didn't miss the cut maybe we
1:08:24
should just bang it up to 420 or something like that keep with the cult number and are there 420 interesting
1:08:29
people in the echo geez they just start clicking follow like
1:08:35
you know anyway back to the poll so we we did a poll and we were i was i was
1:08:42
just probing for what like you know what people appreciate in the show but now i think the polls just more of a
1:08:49
meme because i don't think we actually got the answer that is the real answer
1:08:55
because we've had more uh people watching today than what we ever have had previously right
1:09:01
so like on the live show so obviously people want to see devs but in our bloody pole the options were
1:09:09
dps [ __ ] talk validator talk my ignorance of movies and dev guests
1:09:15
and they were voted pretty much in that order in a lineal type thing so
1:09:22
i mean people appreciate the [ __ ] talk about the dev talk apparently but i think that's [ __ ] because
1:09:29
ethan came on and people showed up for it so well it could also
1:09:34
be that the twitter poll made people believe that there'd be more drama than there actually was so really it was it
1:09:41
was false advertisement really well i mean that that's entirely true so we've really
1:09:46
just rubbed people from the pole of going well you know we're going to have more dps now because that's what you [ __ ] want but
1:09:54
i mean to be to be more scientific your question was what's your favorite part so people showed up
1:10:01
people showing up does that mean it's their favorite we're we're not good at dev interviews so like we do okay with it but we usually
1:10:07
get off rails everything else so the question was what's your favorite part and the favorite part is us doing this but doesn't mean doesn't
1:10:13
necessarily mean that drives more viewers so so i feel like on the dev
1:10:18
side of things i feel like we don't really invite people in here to interview them about their thing right
1:10:24
so like if we have you know when we've had people on in the past like shane and and you know other
1:10:30
other people right devs we haven't really like resolved to be like all right we're
1:10:36
going to talk about your thing and and you know we want to know everything about the thing and chill your thing
1:10:43
it's more like you know we just get on and have a chat and i think that's like better for us
1:10:48
because you know there's other people that do that to like chill people's projects i
1:10:53
just like to talk to the devs and and you know pick their brain about [ __ ] that's been floating around for a while
1:11:00
well i think that's the funny part about it is almost every single time we have people on and like
1:11:06
you watch them right in the first five minutes we're just talking [ __ ] and then 10 minutes rolls by and then 15 minutes
1:11:12
and they're like are we gonna get like straight with questions or what's happening here and then we're like oh we should probably like
1:11:17
bring our guests in for conversation and then we should have a conversation with them rather than like come with questions as they're prepared for
1:11:24
so i think it benefits us to some extent yeah and sometimes we're just trying to figure out what the thing is like i
1:11:29
thought like last week with the skip team that was really good we we talked a lot about skip and met and meb just because we were also learning which i
1:11:36
thought was cool and the same thing back when we talked about with like the graph protocol and things like that because there's a bunch of things that haven't
1:11:42
happened but ethan like i mean like he's the godfather was him right so that was less about trying to understand
1:11:48
it and more just trying to understand who he is which i think is also not bad
1:11:53
uh sorry sorry sorry sir no i was gonna say i can't get past this
1:11:59
haircut it's just like bobbing around and it's it's like it's it keeps
1:12:04
touching the top up there it's it's really sharp
1:12:09
he's not listening he's not even listening i'm gonna just i i'm just gonna i'm just gonna offend us
1:12:15
here and say i think actually like our interviews have been getting better over time like i think
1:12:21
particularly with skip and stuff it's about like um [Music]
1:12:26
so one of the many things i do for money is i actually interview bands for a living and i've done it for nearly
1:12:32
15 years um so doxxed but i i spent you know
1:12:40
the first portion of time i was doing that with like here is my list of questions the questions are all worked
1:12:45
out i had in advance you know no deviation blah blah blah and
1:12:50
you don't get as good answers as when you have some topic areas if you have like really specific things you want to
1:12:56
know like you know with ethan for example like what's your background that's that's an interesting thing like is somebody a low level
1:13:03
dev that you know were they a cs professor or something you know like
1:13:08
in this space you really never know especially with people doing low level work so yeah that's the point where you're like okay that's really specific
1:13:14
question we want to know about but then like you know the rest of it you can kind of let it flow to some
1:13:20
extent and just sort of guide it around when somebody gets to an area where they're sort of adjacent to something
1:13:26
you might want to talk about like the contracts thing you can kind of push them in the direction of saying oh it's interesting you said that because i'm
1:13:32
interested to hear what you have to say about t wasn't let's say um [Music]
1:13:37
and organically that's what i found gets better responses um on the whole is having a set of topic
1:13:44
areas that you can or can't use maybe you don't get to all of them but if you see like an opportunity to guide
1:13:51
the conversation then then it's quite often good to just like push your guests into
1:13:57
having a little bit more focused like just something to kind of in their head to riff off of i think because i think
1:14:03
that's the difficulty when you're you feel like you're on the spot as somebody talking you
1:14:08
it i mean obviously i would always do this anyway because i'm very bad at staying on topic but
1:14:13
i imagine it's also the case that you're having to fill space and it helps to have like somebody sort
1:14:19
of guiding you on to um you know facilitating the conversation
1:14:25
if you like i think it's a fair take i think more of what i was trying to say was if you look
1:14:30
at other interviews that happen i think that i think the question of interview or the same interview is really the key
1:14:36
um value of trying to define here right keyword we're trying to define and i don't think that our cell is that
1:14:42
we're trying to interview we're trying to facilitate facilitate the conversation
1:14:48
yeah that's true that's definitely true i mean it's but i think that that is i you know
1:14:54
pipe up in the comments but i think like how this we you know we joke about how this started was this is just null
1:15:00
weekly harassment call made into a podcast but i think it's also expanded out into like
1:15:06
yeah i think we're naturally kind of quite curious about the space we work in so why not invite people into the tent
1:15:13
you know to shoot the [ __ ] and and because we have you know we're as curious as anybody else and
1:15:19
you know it should be pretty obvious by now we don't [ __ ] know anything so there's also
1:15:26
that should be a tagline yeah so there's some context around that like the our tree house right so
1:15:33
um back back when i started in what was saying like february march last year right in in 21
1:15:41
like i didn't know much and i just used to harass people on text and
1:15:48
then just call them out of the blue and then have like a three hour conversation with you know one of the devs of a chain or
1:15:54
something and and then i come across this motley crew and then you know i would just call the fray on a
1:16:00
saturday morning or my saturday morning was like his friday night and chat for like two hours or something i was
1:16:06
getting extra hot coffee for my parents or something like that and it was just the thing i used to i
1:16:11
just used to call people because you know i'm not the best typer and uh you know annoy the [ __ ] out of people
1:16:17
and and graciously they would talk to me i don't know whether they wanted to or not but i
1:16:24
had some very long productive conversations i learned lots of [ __ ] and so at some point yeah like like the
1:16:31
phrase says we just decided well [ __ ] it let's just live stream that and uh call it some sort of show
1:16:39
and i don't know it kind of feels i don't know if like we should add a format to it but i think it's worth it
1:16:46
to try because whilst um you know
1:16:52
there's the spirit of the show but then there's also like i want people to watch it because
1:16:58
we talk about some interesting [ __ ] and you know whether that's useful to users
1:17:04
or other validators or you know even bringing people into
1:17:09
the ecosystem just through [ __ ] talk i mean i think that's kind of important in itself so i think we can kill two
1:17:16
birds with one stone but i don't you know if it gets to a point where it's just this formatted thing and we're like
1:17:23
you know ticking boxes i think that might become a bit of [ __ ] and we might just have to get back to
1:17:28
how we do it but if it works and and we can get some good conversation out of people um and before we rug them then i
1:17:36
think that's worth it i guess you're getting two rugs for the price of one per episode as well with
1:17:41
the new format right well i don't know maybe now we'll go okay thanks for coming uh please delegate to us and uh
1:17:48
we'll see you next week i don't like that rugged guess is a little bit much i think it's nice okay
1:17:53
thanks and then we could run each other i i i kind of liked how
1:17:59
spicy that was back there yeah i've got a good sentence i i think
1:18:05
that's i think that's kind of cool because i the flip side of this right is that we also choose i guess we don't have people on that don't have a sense
1:18:12
of humor as well um so we've got we've got larry um next week as well
1:18:18
who maybe doesn't know they're gonna get rugged but i think we'd take it with a plum and i
1:18:25
think most of our previous guests would as well um i think i think we've already got people
1:18:30
in the chat they're very keen about the rugging this is you know if you'd had the rugging on your pole i think the
1:18:36
rugging would have been the the number one thing yeah well it's my favorite thing about the rugging is
1:18:42
when people talk about it on twitter like it's a very niche group of people that reference it on twitter
1:18:48
but when then random people see that and they're like oh my god which project rugged you
1:18:57
they're talking about a podcast with a bunch of validators it's it's it's totally fine it's totally fine
1:19:03
uh yeah it is it kind of hurts a little bit to me that we don't get the viewership of other things
1:19:10
i mean like i said does it hurt you i i don't i think i lay awake at night thinking just why
1:19:15
won't people watch us are we that ugly i got a haircut because we don't have one
1:19:30
our content isn't related so i think this is back to you know to talk back to the guest we had on today ether made a very good
1:19:36
point about users right which is that we need more users and ecosystem we need
1:19:42
more you know we have enough devs we don't have enough devs we need more devs but we also need more users and we need
1:19:47
users not investors which is the thing that i i've said
1:19:54
numerous times i think in our private chat um i i know schultzy and i were having this conversation like literally
1:19:59
last week where it's like until the amount you make you can make from just like buying
1:20:06
atom and sitting on it from five years goes down then
1:20:12
you're not going to have users you're only going to have investors and that's i think the thing we i don't think any of us have any
1:20:18
interest in that for the podcast and that's where the numbers are right now you know
1:20:24
ironically this is the place where users should come for their [ __ ] alpha because well yeah we know more [ __ ] than
1:20:31
the regular people right we know we're involved with all of the upcoming projects we've got access to all of
1:20:37
theirs we we just talk to them and ask them questions that isn't public knowledge and then
1:20:42
probably just yeah but is there any connection between good project and and money there really
1:20:48
isn't like it's all animal spirits i think putt most men should mention john maynard keynes in the in the chat
1:20:54
earlier so you know animal spirits animal spirits that's all there is
1:20:59
well i think that it's also like the time we started is significant right like we started
1:21:04
basically with prop 16 and there's been a mass exit of users ever since right
1:21:10
and we were here before 16 [ __ ] me no i know but that was like episode like four
1:21:17
right so we kind of started in that time when there was the exodus and if you look at youtube crypto youtubers right
1:21:22
now a lot of them have actually just kind of like segued into just general news broadcasts like coin bureau coin bureau
1:21:30
does a lot of just kind of general news things right now whatever he was all about just looking into things or
1:21:35
looking into cryptocurrencies and whatever so i think that it's just kind of a weird time in general um and then
1:21:42
yeah of course we should probably do some more marketing and we could tell more people yeah [ __ ] news
1:21:49
number two number two on your poll was validator talk which we honestly we've been kind of ignoring a little bit like we should be bitching more about about
1:21:56
operations and some of these alpha chains we were just talking about because i mean i think there's a hugely
1:22:04
sorry the there is like a hugely you so user and i've been working on are we are we allowed to talk about
1:22:11
this you said i shall see about what about i don't think
1:22:17
we can say what network and we've been on a test now right that we've been cheating a little bit is
1:22:23
that what you mean no but we we can just say we just say like we've been doing some stuff yeah
1:22:28
yeah yeah throw a test now on aptos right and that has been i think that has absorbed all of my
1:22:34
validator brain for the last four weeks i've been doing i've been doing actually quite a bit of routine maintenance on nodes um
1:22:41
we've been doing a bunch of stuff in our aws account um
1:22:46
but that has been very interesting and i think because we we don't we only cover
1:22:52
the cosmos that we i think maybe haven't been talking about validator stuff because
1:22:57
i i think at least for the three of us we've been kind of very very focused on how
1:23:03
insanely different that is yeah and trying to get our head around that and then you kind of come back to cosmos
1:23:09
nodes when you have to do stuff and you're like ah how simple how excellent i just have to do all the things that i
1:23:14
know how to do manageable yeah exactly yeah like doing the uni stuff this week like so we're we're june v10 coming very
1:23:21
soon um look out for a gov prop on a chain near you probably in the next uh 10 days
1:23:27
uh maybe even sooner um and that's been like really like just coming back to that and going like oh
1:23:33
yeah all the node ops i can literally do blindfolded uh has been really
1:23:39
i think that's the difference there isn't because we talked about kajira and stuff where there are there are now chains that have
1:23:44
oracles and whatnot you have to run that being a validator is getting more complicated on some chains but
1:23:50
sure for some others it is still just a case of maintaining high up time maintaining a good quality node rotating
1:23:56
things out you know i think there's been a lot more stalls and that stalls for
1:24:02
non-experienced i mean honestly we talked about how long some of us have been this has been under two years right so so i've had a lot of stalls lately
1:24:10
yeah and so and with the with that happening and you have a lack of experience or just people moving too quick or whatever else i mean
1:24:17
we've seen the number of like the number of tombstones have been gone like crazy um i think that happened like once i
1:24:23
remember in 2021 i remember i remember that one or two times and now it seems like there's like somebody got tombstone last night or something yeah there's
1:24:28
like one on the hub wasn't there in like the entire year yeah something like that yeah and then when you have like block
1:24:34
daemon doing two on atmos and then on the test nets you've seen them all the time and then there's 19 on the same
1:24:39
test net or something like that it's like a parade uh so anyway i think i mean there's
1:24:45
still some interesting there because you know we talked this before but the number of validators and people get involved is is way up but the chains are
1:24:51
pushing uh a lot of change and so that might create stalls and issues and then you have kind of like this gap between
1:24:57
how fast we're moving and maybe the common understanding of what's going on in some of these situations if they're
1:25:03
just running a single node and not you know using some sort of off machine signer or something like that
1:25:09
so yeah it's good times and speaking of shitty chains that are changing uh the merge is tonight right i
1:25:15
hope you guys see anybody running i was gonna say somebody was asking me earlier just what do you think of the merge
1:25:20
because then you're waiting crypto and i was like yeah i only hold a couple of eth so i i guess if it makes if chain go
1:25:27
bur and i make some money i'm happy it's yeah i'm like who [ __ ] cares
1:25:33
we do i think she'll say we want some get we both run some get notes for some i know you run them for um
1:25:39
uh sommelier and gravity and umi that's right yeah so which all which require
1:25:44
that bridge type of structure and so that that it was supposed to be like now it was i think it was scheduled i
1:25:49
remember it was a great time it was like 7 00 pm or 8 p.m tonight or something like that and now it's scheduled at 5 00 a.m utc
1:25:56
um so anyway i'm not sure how much that's when i wake up tomorrow morning to go to
1:26:02
london basically yeah it's gonna happen around there and the hash rate's going to [ __ ] right because i'm not sure where
1:26:07
the hash rate's going people turning things off or something i don't know what that means but if it's that simple but the hash rates go into
1:26:14
hell which means it just keeps so i hope it keeps going down because then it could push out to maybe the morning would be great well i mean i imagine i
1:26:19
think the idea is there are a bunch of miners that are having an
1:26:24
exodus right they're just getting piss off i'm going to what's it called proof of work youth now like like howie
1:26:31
or something like that yeah and so there's quite literally being an exodus of
1:26:36
energy then which is changing comes down and then of course um is from what's it
1:26:41
called the [Music] the death mark or whatever how east created a a point where they had to improve his
1:26:48
state girls the rewards would be so low that might have shut down anyway yeah yeah so have they have they um
1:26:56
did they just fork it into power that's their plan yeah
1:27:02
yeah yeah and then like it's interesting on each side because we have to run a uh
1:27:08
it's a beacon node um so since it's moving to pos and i guess since there will be
1:27:14
the possibility of tremendous number of forks of the chain that now there's these beacon nodes that
1:27:20
are i think helping the actual consensus node figure out hey this is the one that you're following like follow this one uh to try to get
1:27:27
past the merge because i think it's yeah i think i think that's why i understand it but i think the idea is that it's pretty easy for this to fork um into a
1:27:34
uh a bunch of smaller chains and and get some issues going on here so so all that's running um but yeah looks like
1:27:40
5am or some [ __ ] to bring it back to the cosmos there's actually already a proposal
1:27:46
connecting proof of stakeh to ibc out using uh ck stars nice so that's pretty
1:27:52
cool i just tweeted about it um yesterday uh jack zamplin is
1:27:58
on the advisory board
1:28:05
makes me curious about what what does this mean for for gravity or axle r right because suddenly
1:28:11
they kind of are relevant if there's if ibc is already connecting why why do those need to exist anymore
1:28:18
right yeah why do why do you need that kind of a little bit clunky bridge structure right with uh intermediate walls and all
1:28:24
this kind of stuff yeah well i guess you don't if if uh if heath connects then
1:28:31
see you later bridges right if it works out yeah i mean i i for one
1:28:37
and for that i like i hate to say you know let's unemploy early i mean i i the idea would be gravity justin from
1:28:44
gravity would go do something else but it does kind of this kind of unfortunate yeah
1:28:50
just yeah um so it's an unfortunate reality of a
1:28:59
i actually have a question for everybody um with regards to like you know stable coin right so
1:29:06
there needs to be i think there needs to be a stable coin that is
1:29:12
not a [ __ ] algorithmic stable coin like a proper you know centralized
1:29:17
dollar-backed um stable coin in the cosmos at some point right
1:29:23
so um you know nothing against the the uh
1:29:28
you know algorithmic stable coins but really the only one that is really pegged to a us dollar is the ones that
1:29:34
are actually backed by a us dollar that said do you guys think that it
1:29:39
should be a cw20 or an ibc token with its own chain
1:29:54
[Music] you