week [Music]
hello and welcome to game of nodes a weekly podcast on the cosmos from independent valid validator teams Valor
doodle teams that's what I nearly said uh and despite that being a bit of a downer intro as I try to find the
[ __ ] button to press oops demonetized already let's go let's [ __ ] go if you
monetized immediately within the first 10 seconds we have probably the most exciting show that we've we've all been
looking forward to this for like a month or two but working out everybody's schedule has taken [ __ ] ages so we've
got Sunny Jake and Ethan to talk about mesh security which is super [ __ ]
exciting if you're into well if you're into Cosmos making it then you
should be excited about mesh security because otherwise we probably ngmi so
let's go mesh security um right we're just gonna like ask you
some questions about mesh security is that cool or do you guys want to like say something about it first are you like
I've just got to get this off my chest about mesh security what a lot of people don't understand mesh security is is
there anything you want to say before we just ask you dumb questions I think we should introduce it because I
actually told some people to check in they said today we're talking a future proof of stake so like maybe you should
assume some people don't even know what my security is they weren't around the hack was them they have not been following this so make sure you catch
them up first before you go and drop the crazy knowledge okay now you've said you've told people
they're gonna learn about the future of proof of stake it is
um so I mean I started I'll refrain this way I started Cosmos
in 2016 when I got like built both tenement POS I got into that one I heard that like 715 it got like super stoked
about that one and then when it explained IBC you know blockchain communication and how to make Sovereign connections of that this is 2016 right
and this is like mind blown this is next level stuff right you took it okay five years you get everything out four or
five years you get everything implemented but like it took people like three years even realize this was a
thing right like it was so far Advanced that it was like in 2020 we're like oh yeah multi-chain Bridges is a good idea and
you're like yeah I've been working that for like three years guys um so for me my security is one of these next like mind-blowing things right I
don't think it'll take so long to get out there because people are in it but like okay we can have proof of stake okay we dig that we have Bridges like
even IBC secure bridge is important right it's like great what are you building it send tokens you're like no we can build something more right so
people talk about building like you know cross-chain calling composition and we're still working on stuff like that and for me like we're working on that
stuff and then message here is like let's just throw a bomb down here let's like take proof of mistake itself run
that over IBC and have like the like meta proof of stake like level of like like whoa whoa whoa this is what the
hell is going on here right like so I mean this is one of those mind-blowing things when we saw science talk and Cosmo verse I was like whoa like yes yes
this is crazy this is awesome like it might not happen next week and the major
change will take a while to adopt a major change but like yeah let's do it and let's get some tests going on and like convince people like this is
feasible it's realistic and I think there's a lot of naysayers are saying this is too much I can't understand it it's probably wrong but um but man it
was [ __ ] intense and so like here this things have been weren't around uh we had Cosmo versus Medellin uh was that
like a month ago six weeks ago five weeks ago something crazy and uh suddenly dropped the first day that is
crazy talk on my security how's the future and maybe you can give some more explanation I'm saying why it's so awesome and then the hack was I mean
Arctic out there Jake and I didn't sleep for two days um code like Mad Men and ended up with a
[ __ ] like MVP that like it worked it the rewards aren't working but you could stake tokens in one chain it sends over
across a chain and it'll actually stick across take it and like add validators you could take across the valley on the
chain and that was pretty [ __ ] crazy and it's like you know now open source about contributors and um we even got a
record out of it oh my God that's awesome uh thank you
for that one thank you I think that one so um yeah yeah it's good so I I just want to say that framiently this way for
me this is like in terms of infrastructure levels a lot is the biggest mind blow I've had since IBC so um maybe Sunny can explain a
little more like what it is and before I do like the future how's it going out so sunny it's my thing to always ask if
you've ever you I don't think we've ever done any engineering work together so you've probably never seen me do this but I always just say okay everybody
can you just explain this like on five so that's probably another good introduction to like from your point of
view like mesh security Eli five it to me I am a [ __ ]
come on idiot hurry or idiot in a medium level of hurry
um 10 or 15. how about that we can compromise at 10. we can
compromise again so there's this thing called proof of stake and the ideas
what is previous no kidding okay proof of stake it's a way of securing
blockchains and the idea is you people are putting some coin with economic
value to secure a blockchain so you know ethereum is now running proof of stake
it has eat as its collaterals as the staking coin osmosis has osmo Juno has
Juno Cosmos Hub has Adam each of these chains have like and and
the amount of Economic Security each change proof of stake system has is about the value of the coins that are
staked on their chain and this is rough you know you're roughly proportional to the market cap of these coins and so you
know every other blockchain ecosystem has this like idea that like Hey we're
gonna have like one base coin and base chain at the center of everything and it
provides security for everything else built on top of it and on its ecosystem so you know this originally started with
the model of like hey we'll have one chain like ethereum and just all the applications get built on top of it
um and that one eth token will secure everything all the contracts built on top of it we realize over time or they
realize over time we knew that for a while that that's definitely not going to scale uh long term and so then
they're sort of then that move towards this world of like sharding and like Roll-Ups and stuff where it's saying
like hey we can have still have one base chain and have all these other chains but they still get their security from
the one uh base token so this is kind of like what the polka dot model did where they're like hey we have all of these
parent chains uh people can build their new chains but they all have to use dot
staking and use the polka dot validator set in Cosmos we take a different approach
where our goal is to make something that has no single central point of failure
no single token or chain at the center of it but really build this network of sovereign
inter-operating change which is what we've built so far today but you know one of the big questions
that people from other ecosystems always ask is how are you going to secure How
These Chains going to be secure enough you know like osmosis market cap and Juno market cap and stargaze market cap
each of these individually is like way lower than like something like ethereum
or let's say Solana or these like high L1 market cap things and this was driven
largely by this like fact protocol thesis that has been like around in the crypto space so the last like I don't
know five years five or six years which has said that like hey all these l1s are going to get value from the applications
built on top and what we're starting to see is in Cosmos is like applications
are like wait a second why are we giving all this value to these like L ones that were built on top of instead let's just
go build our own app chains but now it's like how are these app games going to secure Each Other Well the answer is
they're all gonna you know build a coalition and secure each other right so
if you look at uh the analogy that we gave in uh I gave him the talk at causal
verse is if you know the L1 normal L1 model of Securities like
this Empire model where it has all of these colonies but you still have one Empire that's like giving security to
everyone else mesh security is more like a NATO model where it's like hey you have a bunch of
sovereign countries that all are securing each other they you know they all have their own militaries they they
have their own independent governance systems but they have a mutual defense pact if
any one of them is attacked they sort of all uh agree to rush to each other's
defense um you know maybe Article 5 will be tested sometime in this coming week we'll see
how that's going but yeah um but so but that that's the
fundamental like base premise and so uh yeah that's kind of the mesh security is
just this idea that hey how can we have many chains that like use the sum of
their economic value to secure all of them rather than one chain only one
token securing one chain uh now the actual implementation of how to do that
there's actually one I think maybe one thing that I feel didn't come across super clearly from the talk is that
there's actually many ways to message security is just this idea and there's many ways of implementing mesh security
there's you know right so I guess like a follow-up question was going to be you know I guess a lot of people will be
familiar with ICS as a concept so I've been talked a lot about a lot there's recently been the um
I am going to get the name wrong because it's very close to the name of this podcast the game of chains recently
right which is the ICS validator uh competition
um and my my understanding correct me if I'm
wrong is that the kind of the jumping off point for mesh is sort of the
ballpark design for where ICS was going for like not version two but potentially
version three is that right
because I I think I this was going back a few months but I think I posted something very facetious on uh Twitter
about I don't really [ __ ] see the point of ICS well somebody you know it's
[ __ ] somebody changed my mind kind of you know typically typically it wasn't quite as rude as that obviously I'm not
actually I'm actually not I'm not a monster anyway um but I think it was Henry from
penumbra who who finally was just like no no you need to think about layered security as the fundamental use case for
this like yeah if it's just like one app chain the argument holds fair enough it probably doesn't do a lot but the
layered security is the the game changer and that that is my understanding then of like sort of where we're going with
mesh which is a really exciting idea right so is it but then so the question would be I think for a lot of people
looking at this so what is the difference then between mesh and and ICS or icsv V3 I suppose
yeah so ICS V3 was this idea that like hey chains can have their own Sovereign
but then their security you can get augmented by you know another chain also
uh you know we call that we've got we've started calling this term cross staking on it mesh security is and and so the
ICS V3 design was saying like hey there'll be one chain that's like you know doing this augmented security for
everyone else mesh security was basically just the idea that hey we can take this mechanism and do a couple of
things things one run it bi-directionally so you can say hey chain B is getting security from chain a
but chain a can also be getting security from chain B that's how you so in it's like weird because in the ICS V3 like
version it's like only one chain is getting the sum of both of the market caps well the other chain the provider
chain is not getting that so that's correct and then the other thing is saying that hey you can actually be
getting so you can be a consumer from multiple providers you don't have to be a consumer of only one provider one's uh
security provider at the same time and that's how you can get this like mesh going so instead of just being a
bilateral thing between two chains you have it's a multilateral thing between many chains
right okay so then I guess the the so the interesting thing uh about IBC is
how it's quite a small protocol like it's quite easy to conceptually understand it's very it's very tightly defined and
and essentially anything that can do the send receive acknowledgment cycle correctly formatted
can talk IBC and kind of heterogeneous um protocols can use it right so it's
actually a way to bridge from Beyond the the cosmiverse if you like
to other ecosystems that might want to bring assets into the cosmos they can be a part of the IBC verse even if they're
not a part of the cosmos ecosystem right so
and this is maybe this is maybe a characteristic me comment of jumping ahead
looking at a cool thing and being like gosh I wonder where that's going in five years what my question is like because a
lot of this is focused on the idea of shared security via dpos is there a way of generalizing the
concept of mesh security further such that you can accommodate potentially
other chains that aren't using dpos as their mechanism but still want to rent security say because you I guess you you
would potentially lose that bi-directional element if the other chain Works in some very fundamentally different way right
I think the bi-directional thing is not core um I think the potential by direction is
core but when you designed it what is it's unirectional but both sides can do unidirectional basically you have a
channel that says chain a can provide a chain B and you can turn that channel around but you just run one of them right you could say that basically you
know hey strategies will take security osmosis or Moses doesn't want to start your security right like totally fine
you can run it both ways you can run it where you go one way and they take four percent of security from osmosis it must
take three percent from stargaze or five right you can run it so they um do both the equal amount they take
there's lots of ways so basically the fundamental unit is this one directional flow and maybe break down the flow and
so I can imagine what this basically saying is um you lock your tokens on the provider
chain right now those can also be used to stake on the provider chain as well like a staking contract you can also
like lock this thing and it'll also stake locally right but if it's not a that's ideally if it's DPS not DPS
doesn't matter as long as you stick it right you could um I mean I guess that most of them are but you could you could say dot you could
stake uh right obviously it dot it's connected somehow right let's say you just take your dot on it you take your
salon on it which is a totally different thing well I guess I guess your point is that as long as it's bridged across it's
the same as the voucher token received over IBC isn't it to some extent so as long as it's locked up on the provider
it's locked up on the provider chain right it's locked yeah just like voucher and what he says vouchers it has multiple so this rather than sending it
here lock them it's going to one chain we lock it here saying this is my collateral here and I will put my
collateral here on a b and c and this value here is probably here this value here right and you said there's vouchers
basically of like I am guaranteeing that if this value in this chain misbehaves I
will be slashed right right and the way you can safely do it that's basically all you're promising the receiving
change okay cool you I trust you provider train to some level right I trust you this one percent and you've
told me you have 300 million dollars you know split among these validators any of these values misbehave that I
said I will tell you this and you'll slash on your side right so you don't move the money you just move the promises basically it allows you to then
promise multiple times this is not until first responses hate is kind of these overcloud realized ponzi's things like
it's not because basically it's like an insurance uh taking the insurance thing that if
you are insuring a thousand cars you're not supposed to take a thousand cars value there's a hundred thousand dollars
and the bank is collateral to prove that you cover him you probably have the money for cover 20 cars in the bank right and maybe you have more big
Pockets later can pull stuff out but right away you can be able to cover 20 of them you don't need to lock up all of those
because the chances that the Thousand are all going to just go wrong at the same time and proof of mistake only
slashes like five or ten percent usually so like now if your money is here and some three different chains even if they
all slash you you won't even lose all your money so I guess the other interesting thing here is like you know
you could have as long as there's a mechanism for lock up which could as well be a smart contract rather than say
A validator you could have say like a proof of work chain um also in that mix as long as it talks
IBC so from what you're saying you can see mesh as just a pure addition to IBC
because the provider chain only needs to conform just like it would to IBC it has
to conform to that protocol it just has to conform to a mechanism implemented in
their language in their way of doing things that will ultimately essentially collateralize that thing in
the way that it says it will right and then you then and then the party can continue so that's that's like I think quite an
exciting idea in terms of like because the great strength of IBC is that if dpos doesn't make it if it's
ngmi on the longer term IBC may be GMI anyway right and so that's the great
strength of focusing on protocols rather than a specific technology and like that was the bit that was maybe less clear in
my head about mesh security but from that description it sounds like it's still very much in the same
spiritual School of Cosmos design I guess where it's sort of protocol protocol first we've had a couple of
questions in the chat so um if you don't mind I'll just I'm just going to bring those uh bring those up so Luis Quezada
has said how different is mesh from what Celestia is trying to do is it not related at all
yeah um so Mass security is doing something
pretty different than what Celestia is doing so let Celestia is trying to
solve something called Data availability which is a way of making
uh sure that the data of a blockchain is actually
widely distributed and not in on Beyond just the validator set themselves
because currently you know what you could happen What could happen is if a
validator set goes malicious they could just like make blocks but not share the
data of the blocks with anyone else and they only have to send the headers to other chains to get their IBC uh working
there's reasons why data availability is important uh for you know I think
my heart take is I think people kind of a little bit over emphasize the importance of data availability especially in the cosmos architecture
but we can go and I can talk a little bit about that uh mesh security is
mostly for solving um you know increasing the economic security of slashable things so one
thing with data availability is it's not very slashable there's no ways of creating proofs of data availability
that availability today so security is focused on solving uh you know making it
solving the state primarily focused on the safety problems you know I think you
can eventually do like liveness slashing and stuff over uh mesh security as well
but I think the primary focus is solving the safety thing so the first thing is just make like doing stopping double
signs right so you know that's something that we already know how to prove very easily today
um and create proofs for but eventually we'll have more sorts of evidence that is slashable for so uh for example
um uh like once we have like validity proofs uh or like you know it's not all
Cosmos chains are creating like snarks that can be verified you can use those as a slap you know if if some if a
validator makes a state transition that is invalid you can have that be a slashable condition as well so
uh and so maybe one day if we have like data availability like fraud proofs or whatever that looks like that could also
be you know used with uh mesh security to increase the economic uh weight
that's backing these uh the Mass security guarantees okay so and then following on from that
we've had we've had a question in the chat about um slashing which are uh kind of bring
up so Adam Bomb says say say I have 10 atom cross state across n chains is
there a limit of change that I can cross stake to like a maximum value for n
um does my total slash potential have to stay below 10 atom that's a very good question I think we
have to differentiate my security from the Mass security IVC protocol and the security implementation as is and the
ideal my security implementation I have is MVP right so currently we have this like proof of concept which you've improved a little bit it does something
um and that one it doesn't matter you can have N I didn't care right it doesn't care at all and the security
protocol also doesn't really care it basically says I trust you provider
chain that you will honor this slash if I slash you're on the validator you will slash those tokens right so I'm not
telling you how to guarantee you can do it I'm gonna tell you how to guarantee I just want you to guarantee me that you
will do it right you give me a promise that you have 300 million dollars of their tokens locked up on your side
about 300 million x tokens locked inside and you promise you will slash them at 10 if I give you a double silent proof
right like that's what your promise is so the packet rather than saying ICD-10 token I promise to release 10 tokens if
you give me the voucher for IBC ics20 right I send you this telling saying I have 10 tokens locked up they'll release
whoever's name it's a promise of I have this many tokens locked up they're willing to do at this unbonding period
with this slashing Edition right that's basically what you promised right that's a packet level you're giving this promise that in slash
um then the implementation can do what we do now is we don't care or we can say
no we just do that we have one lockup that says okay you want five percent you want ten percent you want 20 okay and
there's some can't be over there so if all the all these places slash you that you will be able to cover them right we
couldn't force that and I think that's actually good idea it's not required it's definitely not done yet it might be
for MVP and someone else can Implement differently but I think it would be a good idea to add it sure and it's totally like implementation dependent
you're just supposed to say that I guarantee implementation if you're a provider and you're allowing people at the overclockwise 100 times
if that got their thing right then people might not want to accept your provider tokens anymore
right so but then there's also a question which I think maybe came up in Sonny's talk I can't remember
um at that cosmiverse which is uh or maybe it was actually in the notes for the original
um MVP that you guys did the hack which is like there's kind of this outstanding question about token value normalization
which is strikes me as like you've you've kind of uh it's like a
hard problem that's like a oh suddenly we also kind of need to understand how
we're agreeing at a definition of value at the point of lock-up right
um and then and then which and and you know then and then we're sort of into like the murky world of like well what
is this what is this thing worth and and what what is the ratio that then can be applied to these different weights and
things so then that that then has this interest so it's interesting as well it's here Sunny
talking a minute ago about um you know zero knowledge proofs and stuff like that because there's there's
a question um which is there how how do you bring and I guess maybe this is also
tangentially related to some of the Celestia stuff it's like how do you bring stuff that you know is outside of
consensus into consensus by the way of a proof that's sort of binding or
verifiable in some way right and is that is that a general is that concept always an oracle
or is that concept a zero knowledge proof or something and in addition like
if you're talking about mesh security occurring on every chain or is every chain going to have to go okay well we're going to call an agreed Oracle
chain to find out that information or is every chain going to be sort of running its own Oracle in the future because
again like some of these questions I think have quite a big bearing on the app chain thesis like are we going to be
able to literally you know trivially even to just do something with mash like staking we're just going to assume hey
also in the background you are going to call the Oracle chain okay well if band wasn't [ __ ] you're cool banned and
find that out and then your stake will happen and we normalize the value of the token and we move on and sunshine and
Roses because it seems like it I mean obviously like IBC took like four years to stabilize and we're very early on
this but it's kind of interesting how quickly we start to find General hard
problems under the hood like like that one you know
the same way I said the protocol doesn't care about this the protocol IBC protocol doesn't care about this it
allows Innovation on both sides we talked about the staking side which says okay how many times can you
collateralize your loan I mean over like what level I stock up and lock on three different change how to handle that
that's a promise that collateralize you're supposed to do in front you can put in different ways we can experiment the the that's a provider chain right
the consumer chain basically says I have one million Juno and 500 000 osmo and 40
million stars on three different things supporting me its job is to normalize them into its native token and to figure
out actually limit them and say hey cool I don't care if you put all the osmo there I'm only letting osmo have up to
40 20 my voting power for example right so it's jobs it's the job of the consumer chain the consumer can play any
way it wants to which is basically normalizing remote tokens to its token and then also providing like limits on
percentages uh for that so maybe I'll suddenly give more on the theory behind it but I'm saying this is total
implementation dependent on the cons on the consumer chain and they can choose to do it and they innovate that way without the protocol being the same
thing different people can have different ideas oh this is working this we should add this check here we add a check here and it's totally compatible
with the protocols without requiring the whole state of all the mesh to update setting each update independently and
expect with new things while saying we just give these promises right we honor your tokens and that we honor your slashing
yeah um so this question about like how should change get the Oracle like or the
relative value of different tokens um so the view my view here is that we don't
need a fast Oracle for this what we need is a slow Oracle for this and slow
enough even maybe to the fact that like governance itself can act like this Oracle so you know uh on osmosis we do
like uh incentive changes via governance on a weekly Cadence and I imagine like
just re-updating these some of these parameters could could come through that
um the and the reason you don't need a fast Oracle is because like Ethan just
mentioned like all of the Val the the relatives taking value of these tokens should all have a pretty high discount
applied to them and you have like limits on how much they get so as an example we actually do this with super so one thing
that's worth noting is super fluid staking on osmosis is actually very similar in a lot of ways to this mesh
security design uh where you know what is mesh security it's governance making
decisions on which tokens which other chains show tokens to accept as part of your proof of stake system superfood
staking was basically osmosis governance making decisions of which LP shares of
other people's tokens to accept in our proof-of-stake system and what what
happens is what we do is we apply a 50 discount so we're using our osmosis
on-chain Pool prices but we're providing a 50 we're doing a 50 discount to the uh
value of those tokens and that that allows us that we only update the relative values of these tokens once a
day so even if there is like some fluctuation in those in the values the hope is that there's not going to be
more than a 100 or 50 on upside 50 on the downside change in the relative
value of any one of the tokens like at a faster Cadence and like one day or if
there is you know we should be having like some sort of circuit breaker for those situations anyways right so
um yeah that that's a general idea it's like because we provide enough of a discount factor to the to the staking value it allows us to use a slower
Oracle because it doesn't always need to we don't need to know the exact exchange rate between any of these two tokens
it's more of like a you know rough as long as it's within you know some some
rough uh okay let's say explain like I'm five moment why why is
the discount needed um the discount is needed for one couple
of reasons one you don't well we want these limits to make sure that another chain doesn't like just have complete
control over your chain so you you know you want to cap let's say you're doing mesh security with like 10 other chains
right you don't want any one of them their economic weight you know let's say one of them is
like much bigger than any of the others or they're you know one changes happens to be doing a lot more cross staking
than the others you don't want them to suddenly get like a third of your uh chains voting power so you cap you can
say like Hey we're capping any other individual chain at a maximum of like 10 of our Train's voting power so that way
you know even if one of these tokens goes through a Luna like situation where
the the value of those staking tokens becomes like you know effectively got the chain could have been taken over
then you know it doesn't have this like takeover power on your chain sure so I understand that in terms of
like the the cap of I guess like value or voting power but where where does like the need for a discount come in so
far as it affects the Cadence of needing to attain prices so yeah so so the
discount comes from you know how much you're willing depending on how fast you want your Oracle to update uh that
that's kind of why in like the osmosis case we put a 50 discount because we're like okay we're making this bet that
like it's not going to the price is not going to go up 50 or right so that's so the 50 is more a hedge against potential
volatility than it is like a hard and important not like if you had a super fast Oracle available at like a
millisecond K well not millisecond Cadence when Matic is blocks are still six seconds but five seconds whatever
like you could you could in theory if you had an oracle at the Cadence of your blocks then not have
um a discount and just purely have we want to cap voting power at this and actually we can trust that the price ice
like we're not going to lose 50 of the value between one block and the next it's just [ __ ] impossible right well
actually what you want to what you're actually kind of making a bet on is really it's not going to change that much within the course of an unbonding
period because what you don't want so in the super fluids taking example let's say you have one token that's like worth
you know or let's say a Luna situation where it's worth a lot of money um and then it's like has a high value
in this in our in your staking protocol and then a week later that those LP
Shares are worth basically nothing but at a block less than three weeks ago or
two weeks ago and also like lesson on one period ago it's actually worth a lot more than it was well now it's like hey
you could you you could do some double signs at this older block and you're gonna get slashed at the current block
but that's fine because those LP Shares are worth basically nothing in the current block so that discount factors
also it's not just based off the Oracle update it's also based off of like assumptions on the volatility during an
unbonding period right okay that makes sense uh so I
guess kind of relate a question on on governance uh tiger grades has asked um
does mesh security effect in any way the governance in each chain or does it
remain the same based on token weight held natively I think the answer to that
is you can decide right that's the the notion but I love his opinions but
because Osmos and Juno are probably to the change that will be implementing this I'd love to hear their opinions um there is currently no way for delhi's
to override their vote let's put this right in their protocol we could add this it's interesting India very very good question uh thank you target empty
that's a great question um So currently you just delegate to validate remotely right
um and the current implementation says if you have remote stake invalidated X or local
statement X it doesn't matter right like if that much stake is on them at the end normalizes all the discounts et cetera et cetera
um those count the same as they will be they're that value to get more voting power for terminate and more voting
power for governance right same thing but the remote delegator cannot override the vote on the validator that's your
current impotation um I don't know if we should make it so the remote delegators can override it
where the current position is good or whether you should limit it so we separate out the remote vote actually doesn't increase the tenement power of
those validators but has no impact on the government's power those are all actually three interesting ideas I'd love to hear the opinions of both other
two chains on which one makes sense for them but yes anyone could pick it out except adding remote governance from
delegator I think would act require protocol changes can I uh can I chime in with a question
here um The Fray are we Midway through another another thing is it time is it question
time yet um yeah and look I've missed a couple of
things here because my connection has been horrible it keeps bloody dropping out pretty much exclusively when Sunny starts talking
he's even in your country are you here I'm in Sydney right now by
the way Sunny where the hell are you I don't know if he's talking it's like
literally just Frozen as soon as I said that but um anyway so Melbourne
Melbourne Melbourne um Melbourne Melbourne I mean um I'm in
Sydney so I just I drove down here yesterday that's why I'm in a taji Arts Hotel it's it really sucks
foreign hackathon in Melbourne which is why I'm wearing this uh uh hoodie oh yeah that's
pretty awesome man just whip up here like I'm gonna be here till uh Saturday so you know I'm free all Saturday
um anyway so I was just thinking well you guys are talking and the the information I need to actually
decipher my problems already been probably talked about but um is there any potential for like Ultra
whales on a chain from um they're cross-staking so I don't know
whether we've already talked about whether this is like a liquid staking thing where you can stake your already
staked stuff say Juno then to osmo but say you get someone like
any any validate uh any big validator like um
you know Cosmo station right who obviously have just shitloads of tokens all over the place
um and they you know want to do a proposal say on um
Adam and they really really wanted to pass and they take all of the supported
chains tokens um and still maintaining the price risk on all those external chains but then
just cross stake it all to their validator on um you know another chain is that is
that going to be a risk of like ending up with ultra whales for governance votes
well there's a lot of open questions about the whole governance thing um like do you even want do you even
want to allow it or do we want to separate it like Ethan was talking about for example like you increase your
temperament power but you don't increase governance power which would like prevent such things I think that to me
that sounds sensible I like the idea of like the chains remain like Sovereign and it's ultimately in control of the
people that actually hold the token like you hold Juno or you hold osmo and that's the the token of those chains
um and you know if you're cross-staking your Juno on osmo I'm not sure that should you know get you a vote on osmo
um that that'd be like my kind of taker preference um the there's a lot of room to explore
like mesh kind of governance systems as well it's another whole class of mesh
protocols that could be like very interesting we're still in the very early days but I guess to answer your question
I am personally kind of like I don't think we should like
you know like allow for like cross like we could
build governance into my security protocol and that's actually something that some people might choose to do
um but I'm not sure it's a good idea
so basically to this point that those kinds of things haven't really been
um firmed out yet and you probably need uh you know I guess I'd probably need to
be experimented with in in the um in the test nets for they went to a main net
did you want to say something Ethan was that I think first of all I definitely need to test this stuff out so we're working on a prototype work a functional
MVP that can be done test Nets and maybe some low value chains could try it out but definitely with these attack factors
so you know putting it on uh or Adam as a consumer soon is definitely not you
know let's start with the small ones and consumers um the provider side's more clear right I'm making notes Here good questions
guys I'm taking notes and we'll probably discuss them and come up with some stuff and make issues on the repo um but what I'm saying is this is
actually when Jake said tell me this thing and this is yeah from Sunny the mind blown is it's not about validators
cross taking it's not ICS where values prostate is delegated right it doesn't
matter so Cosmo station can't use tokens from other chain anything else doesn't matter it's the delegator if I delegate
on on the Club to Cosmo station and I delegate to sick on osmosis I can do
that and they have no weight for my Approach it's just me as a delegator locking my tokens and able to delegate
those tokens on multiple Chains It's really the D in dtps right so the valid is just recipient to that so A validator
can benefit from the cross stake atom so if they're a bunch of atom cross technosis that happens to choose
compensation osmosis that will benefit them right but they as a value don't have a way of throwing extra wages doing
it just unless they attract more votes so I guess what I meant was that um you know the some of these really big
validators just have a lot of other tokens in their own Wallets on on other chains
um I guess was more the point um that they could maintain so
can you with the cross staking can you like be delegated on Juno and then delegate the same tokens again on
another chain is one big question you can so they could have all of their
tokens staked across you know all of the chains they they are on they could use the same tokens to stake three times
across all of those chains and then they would increase their vote Power by quite a bit across all of those different
networks as the soft stick the self-stick is is valuable tokens they
can is delegate in that sense so another question I had which is a little bit unrelated to that one
um but probably like something Jake might be able to answer or maybe even The Fray so on Juno uh and this this
sort of follows on from my understanding of um the mesh staking protocol which is
you know not Giga level so feel free to just like correct me a lot here
um with Juno there's the mint module is quite different to other networks in
that it actually switches over to the next phase of um
inflation based on the amount of minted tokens so
if and I'm taking a guess here because I don't think I've actually received this information yet but if cross staking
actually mints tokens um on the consumer chain wait the host chain
whichever side it is um will that then throw like the total uh
amount of minted tokens out of whack and potentially the uh the inflation
schedule change prematurely no the answer is no because
we wouldn't let that happen first of all um second of all uh you know right now uh
the current like implementation does not involve minting tokens um now I believe that our future
implementation wait wait wait wait I believe a future implementation will um but the current implementation has to
be funded by the community pool so the community pool has to vote to you know funds you know rewards effectively for
all the cross stickers so no new tokens minted right now but I think ideally we actually do something that's like super
fluid staking where we um and Sonny can talk more about this um but that's not ready yet and when it
would be ready we would make sure that it works with like the genomenting module because yeah is so I think it
would have been because I remember that um in the hack
wasm um presentation I think he's actually talked about that type of implementation
based on the super fluid staking um situation on osmo so maybe that's
where I got that from so quick quick just follow up from that
the uh the community pool is that being used because government governance has
to approve the incentives or could it be any valid
address with funds that is specified as the correct wallet address by governance
so like in theory could you have like a doubt our multi-sig where you just say oh we'll make a subject let's manage this
and we'll give it funds as long as it's an address that's agreed on by governance you saw it right
okay so so actually that means that almost any chain that has back 32s can
work with this right like well or anyone that's really I
guess that's the point it doesn't have to it doesn't have to have the concept implementation requires cosmosm that's
that's the main requirement okay yeah so the implementation I think once uh you know once we get a little farther with
the Prototype I think in like q1 and be like really great when like the osmosis team has more bandwidth to like have
like a generalized super fluid module like on the on the like native SDK side that can like kind of allow for like a
better user experience around it so um we're actually getting really close with our prototype there's we have like
basic rewards they work we have to refactor them a little bit but it's like it's mostly there
um and then uh going to basically the slasher is the last thing um and then we have a working prototype we can play around with
um but you know I think the the ux won't be as great as it could be if we have like a native SDK module so to make like
a really like what if you already have staked Juno or you already have staked osmo yeah well for this prototype you're
gonna have to unstake them or you're gonna have to have basically liquid tokens in order to use it that's obviously not ideal and so
um you know this is where I think some like development on the SDK side can make like you know cross sticky and like a
really great or make mesh security and participate in mesh security a really great user experience
um we are going to have a UI for it though which is also pretty sick um we have a we have a team working on
that designed out shout out to design Dow um it's gonna be sick
so and again like sort of picking up on this oh we've had we've had a couple of questions in the in the chat about cross
slashing I think Ethan covered that was saying well it's actually up to Consumer
chains to kind of decide how they Implement that against the the the protocol and it's obviously up to
Consumer change to decide how they want to connect to other chains uh you know so do their own research if you like
before agreeing to the other end of the handshake um but does this have like an additional
implication in terms of um so this is kind of a meta layer right
above just running on an app chain at the moment the only real contingent is that you have to be running cosmosm for
the current implementation but what it is doing I mean there is
obviously an idea that it will work in the same way as superfood staking which does require you to have locked assets
on the chain obviously the the the protocol itself is I think as we've discussed broad enough
that as long as the assets do get logged it doesn't have to be staking but there is a question here
about like you're moving the abstraction for the user away from chain level to a meta chain level is that going to
ultimately uh cause problems of say infrastructure provision for smaller
chains if People's Focus moves from sort of you know chain level Focus to trying to
work at a higher level of abstraction because you know the The Logical conclusion from this will be people will
look at it and go ah you know it's actually quite inconvenient that I even have to stake my tokens in the first place so what if I come up with a
derivative way of meaning that I don't have to do that at all I can just operate only at mesh level like it won't be too long before some brain box works
out way of doing that and then that might have uh
uh I guess it's still the thing is that it's still got to be locked somewhere there's still has to be a mechanism for locking even if it's at a meta level right and there's not possible unless
one of the chains at least has some bonded value but is there is there a potential Vector for I guess essentially
creating more Economic Security at the cost of taking away the app chain benefits from
smaller chains that are part of like let's say a cluster of chains because if you're like
let's say you're a new chain and you're in the kind of cluster of chains around osmo Juno right stargaze
if you were just like the fourth chain in that little cluster is there a danger
to your vowel set there I mean obviously you've you can limit the VP so your governance is secure but you're kind of
going to be like stifled by the people around you if you see what I mean like from an app chain perspective it might be counterproductive
yeah so I think I'm not sure if I fully understand the question but it seems to
be like a question of like hey like early stakers and validators tend to be
like the Bedrock on top of which a lot of these changes communities get built and so it's like will mess security like
harm that Community like the usage of proof of stake as it's like Community Building thing because a lot of your
stickers might not be actually your token holders we're actually just becoming from others right they're moving to a higher level of abstraction
than just yeah I am a holder on chain X chain y chain Z they're instead I just
hold some Cosmos tokens I stake them on some chain didn't matter I cross stake them yeah it doesn't really matter which one
because all these all these chains are talking bi-directionally let's imagine that they are in this future world
so I mean that's kind of the whole point of like capping the voting powers and also like capping the rewards that
different like the idea is that like you should be augmenting uh security but
like if you're more bullish bullish on osmosis than anything else the incentive
should still be to stake pure like pure steak osmo rather than cross-staking
something else right because you know I think that's how you get governance power I think you know things like Mev
Revenue might only go to like direct sellers or something like that uh and so there are like there are like reasons
why uh you you don't you would still be incentivized like if there's a new app train that comes along that you're like
really excited about the potential for that app chain specifically there are still reasons why it makes sense to you
know acquire that change token and stake it directly rather than just cross staking
the cross staking is sort of just a nice
you're starting to earn the token of this of this new chain as well and so you know there's this idea that like you
know I think maybe airdrops are going to start to move into that direction as well where instead of like just air
dropping to you know let's say Juno holders you can offer cross taking to
Juno stakers and that's a way for Juno stakers to like earn your new app chains
tokens and you know so yeah I think there's like uh I think I think it will
I think there'll be new ways of community building that takes advantage of mesh security
okay makes sense are you talking about yeah I mean also like oh yeah I mean I was just gonna actually
agree with most of that um I don't think it'll be like much of a problem if anything it'll just Force app
chains to actually build good products that people care about it's not enough anymore to just like launch a cosmos
chain and be like oh here's this airdrop and stake it you know I think uh The Meta is always changing and I think
that's a good thing like build great products and people will care about it awesome
yeah I think the last point in the uh that's anyway and go again on that one so instead of airdropping you basically
often crosstalk it you can have like a crazy reward the first six months right so you give to people that actually built the product right the the actual
coordinators the people that were pre-genesis involved they get a bunch right they can stake their own stuff and anyone wants to jump in
um will come in right and get because they cross taking the odds with a Juno ever to help this thing actually be
secure the first six months of his life he didn't believes in it right and if you believe it you're like oh I wasn't actually involved in this random uh you
know bling chain and but like it sounds cool they launched this thing but I have
a bunch of osmo I'm crossing my osmo to help them support it and I get a whole bunch of bling now now I'm actually supposed to bling right and I only do a
vaccine so rather than doing like an airdrop if you were going to do an airdrop you could say have a vest Cliff
right that starts say six months in the future but launch we're cross-staking enabled and so your initial liquidity
your initial kind of in interest is going to come from the the chains you're in that cluster with
that you have that Communication channel open with which ones you authorize you say okay after three most centers you
know and photos into osmo and 22 stargaze and and I mean
up to that much you guys come in you bring it and so one you buy security from day one you're providing something
right it doesn't cost anything really except your risk slashing risk but you usually took a surprise security and you
bring attention to it right if you care about us now you're exposing yourself hey there's an option to to secure this
other chain and get more or less free at your tokens because you already you know you already lock your tokens for asthma
anyway because you believe in osmo it's like well I'm locked from Osmonds well life is bling thing too and you think it's a stupid coin you want to do it so
it kind of attracts those people and gives them in the show tokens so I think it's a great way of um it's like a better than air dot for me it's called a
mass drop maybe something like I think I might drop maybe guy
give him time at the end just in case
um we don't know what's in that market could be whiskey um Ethan asked uh Ethan asked one of us
to to share the mesh security diagram I'm gonna have to do it as an overlay so
it's going to look a bit weird apologies for those of you who are watching on on YouTube but uh here we go
oh there we go yeah cool so this is I made this a few days ago I was talking
to Jake and put this out there so you can find this on the github.com
mesh Dash security repo check it out um in the video description
um and it's also in the chat yep everywhere it's everywhere so check it out this is basically an image of
everything we're talking about the top is um this image you can look into it might help make sense everything you're saying
in the beginning this is the current implementation and what else are we thinking of it's not the final one there's always details about how the
governance works great questions but basically the top is you lock up osmo on that
a second and top one but then there's native locking and then sends to multiple cross-section multiple chains
they send over IVC the consumer then funnels for multiple chains funnels of multiple change into one meta-staking
thing which is where we want to have you know super fluids taking type module control there so it can do all the
normalization they can figure out what tokens worth what discounts they should apply and do that kind of super fluid
thing to then give those you say okay there's only 10 different validator Power not governance valid whatever else
and that's the one that kind of funnels it in so those the protocol between the green and green is more or less fixed
like the IVC protocol and their functionality is more or less standardized but how the lockup Works in
touch native staking and how The Meta staking one ties the bottom that's huge determination there all these questions about who is governance power where does
it go is there limitations to how much you can do how to discount how to handle oracles those are basically those two purple boxes which I think are really
really good place to experiment either something that's fundamental for a question perspective but is is each
delegator responsible for making decisions on on where they are doing that cross stake or is that everyone
who's delegated to that validator is automatically a part of those chains where they get pulled in it's delicate
Centric maybe suddenly can say a little more yeah because his brilliant idea it's not validating so I take my tokens
I say I'm an osmo I put my osmo on on sick on osmosis and I'm gonna throw it
over there on um I don't know Maiden blocks on whatever because of them and I'll throw it on um I don't know who
else is Target here's your button but not gonna happen and um and I I cross
taken these ones and I make a decision I couldn't read them all the same this delegate e-centric just reusing the delegation so basically think of as a
collateral cool you pull the collateral and then you can put out multiple uh I'm basically ensuring I'm using my same
collateral to ensure multiple parties it's not it's very very different ICS which is I pick one validated the
validation wants the money right right and then I pick a value based on what they're doing the money no it's a very delicate Centric one whereas basically
you opt in to that and there can be can there be a mix of of
um validators who are not part of that IBC structure that are running natively on chain or that's purely
so there's because I know that what's in the ice in The Interchange security kind of road map as well right
um but can there be a mix of that where you might have native validators that are only running you know they're only carrying tokens
locally as well as IBC so and I think you guys are adding yes so the extension of that is that that validator set on
the consumer chain has a VP percentage that VP percentage is a portion of native tokens as well as ibc'd
right mesh security tokens and that is the total VP percentage of that
validator set right and so those add up one to one based on the value in the Oracle structure that we were talking
about before some challenges that are trying to understand what the what that relationship is right that might change per day
and so that validated percentage actually can change very it can change this significantly
not only based on the number of tokens that are take but also by the value of the stake tokens that sit behind that right that are sitting across that
dotted line that you had in there is that is that an accurate summation of some of that people yeah and we we did
talk a little bit earlier about like oracles and fast foods slow oracles we talked about voting power caps so like
you know kind of like mitigate worst case scenarios especially with like extremely large chains
um sure you know and I think we've had like great discussion about like you know what we'd like to see in terms of like governance like you know the actual
like voting on proposals not just like tender mint like voting power um and and everything else works the
same way like blocks get produced based on voting percentage etc etc I know the governance piece aside I can understand the kind of
crosstalk yeah it's actually yeah in many ways I think it's a lot simpler than ICS
and um in some ways because there's like you know no requirement to like run a validator on the other chain or whatever
again it's like breakfast around delegators which kind of I think makes it a lot more elegant and
um you know I think we'll maybe even see like various different like flavors of mesh like for example like you could
imagine a stargazed flavor of mesh security where you can like cross stake your nfts or whatever right
not for a while someday I feel like in comparison of ICS and
mesh that um you know ics's seems to be a protocol
where the um you know the gas chain has to buy security from
the host chain and and mesh seems more like a protocol where delegators can get
paid to provide security to the host chain
yeah the consumer chain offers rewards to any one of these various host
chains that wants to delegate to them right right yeah no permission the poster changes have to
wait last you you say look we're offering money here if you want to and anything that wants to opt in Ops into
it right you don't have to go through governance votes and do some stuff and don't even care the validators and they may say hey the validators on on this
chain hey none of them almost none of them are running on this other chain B doesn't matter because you can cross-date to some other validator so
like you don't recording the whole value set there's all this politics which is kind of like the thing you want to understand security to get away from
like having to go through the politics like I just want security without thinking about it I'll pay for it and so I think this is a much nicer like hey
you just tweet out this money this thing out there's we have rewards for people that's cross stake and if you get
attention you get attention you don't have to go through the Valor set it's basically it's crowdsourcing and if a few people uh opt in they probably get
much higher rewards and a lot of people opt in one chain then they squish out their cap right so they just can't produce everyone's so um I think that
validators I think that validator Centric versus delegator Centric needs to come out stronger I think that's really the that's really the core of it
right um that's one point I think yeah yeah it's a very powerful narrative it is my
whole bitching on ICS on this show for the last since it's been is a good idea
around a voting power structure that exists today which I think is not realistic and and not necessarily great
it gets extended to a bunch of chains that have to inherit that and they have to inherit that from a Governor's perspective they have to hear that from
a VP perspective and I think that's a bunch of [ __ ] honestly so but it's also it's also imposing a hub and smoke
spokes model on an ecosystem that has thrived by not having any kind of a hub and space model but just by having IBC
communicating between connected zones so it's sort of like uh it's a retrograde
step in protocol design yeah which is yeah it seems like very anti-cosmos when I heard about measure
security I was like this is Pokemon 2.0 like it's yeah it's just like my opinion man in the words I mean I'll ask the
host now I'm asking questions to host now what do you guys vote on item 2.0 no comment
I don't actually yeah we don't validate that I don't I don't hold a lot I don't hold any yeah hold on yeah
okay here's a question for the hosts all right put your hand up if you hold Adam
like over time over 100.
I I do not hold over a thousand atom I don't think I hold over a thousand awesome now I am a PO I'm a pleasure
Cosmos Hub Cosmos individual my my life started with osmosis so everything on
that is where I started That's Mike I think ultimately we're alt Maxis
is that one of this okay that's a very polite way of saying total DJ ends yeah
they're incredible total DJs it's a well-known validators are the biggest d-gens in the ecosystem
in this veneer of technical respectability to cover up the fact that
it is just degeneracy all the way down no that's not true I get I get involved in chains that I I think have
fundamental core value you still don't are you still on kerberus spicy
foreign
[Laughter]
two weeks so that was uh that was fun
they just screwed up our billing and they're just like the second you don't pay a second you don't give Jeff a taste
you're done we're just like oh sorry we were trying to just switch to invoicing like that was the thing we were trying
to do in there like we don't care you didn't pay and you're like literally please could you just
can you just like resolve this we're trying to pay you know like you've never even paid Jeff
your protection money we're gonna [ __ ] crush you like a bug and you're like hmm maybe this is why people run
bare metal after all the other thing the other thing coming back to the topic at here the other thing I really love about this idea and
the idea about my security and being delegated focused is that it really evens out the ability for a validator to
communicate to Market to be able to kind of punch above their weight when it
comes to attracting delegators to maybe a new chain or something similar that they might be validating for
um that even though Cosmo station somebody else might have a thousand times as many tokens they don't give a
[ __ ] and and there's a lot of chains that they don't give a [ __ ] about right so it becomes a distraction or something
similar to that and so I I think that's awesome to be able to make it um where they're not just inheriting VP
because where they sit in a validator and a validator set on another chain right so I think that's really cool
we had a question in while we were just talking that from Ben Davis who often asks sensible questions in the chat uh
saying what did you guys make of informal's analysis what's that in reference to have I
missed have I missed some how much uh some Twitter
comments check no yeah they did some and asked this with
like a lot of really big math equations in it and it's kind of funny because like mesh
security is like not even like like we're still working on it so it's like what what did they like actually analyze
um they also released a whole UI to Tool they went to the effort of building a
whole UI tool to just you know as a comparison um
I think you know I'm I I'd be like really down to work with like you know informal as we like move closer to like
a final version of mesh security I think that'd be really great um you know this early analysis I think
is really is really early oh it's a little bit too early we have to like you know finish it
um but I will say I do think they like kind of counted like the economic security
wrong um they looked at the voting power cap and then they like so 300 million is
staked staked on the cosmos Hub but there's a 10 like voting power cap they took only like 10 of that so it was like
30 million of economic value um and I think that's just looking at it the wrong way what's actually there's a
300 million Stakes there or there's 300 million cross stakes and there's 300 million worth of security it's only that
the atom stickers are getting the rewards for 30 million because of the the cap if that makes sense right but
you also get expensive by them basically didn't you because I think we're potential I think
they're potentially arguing now now I've understood that argument or like understood the kind of Crux of where they're coming from on it they're not
arguing in good faith are they because um
like it's funny if you say mystically oh there's a funny idea never happened uh you know Jake and I stay up all night go
too long when you say hey there's something here like oh [ __ ] we have to prove it's not good enough so there's definitely it's fun like you've been
working out for a year and a half it'll prove the largest bed and theirs right and even though it's like it was a prototype it
like they've invested a lot of their time and energy into thinking about informal yeah sure but the analysis I say one point
analysis is besides that other thing is the analysis is largely about one large chain
protecting other change right right and then this is not about four chains of similar weights cross taking each other
right which is really the which is really what messenger is about right they're still thinking about hey if the
Hub is running mesh security there's a hub providing Direct Security look you have more security if you're 100 for the
Hub than if you're only a little which is true it's true if you want to have the Hub security and the Hub is 10 times
bigger than any other chain let's say right the only way that level is if you're 100 goes over the Hub and if you
want to have like if you have but then if the Hub is not 10 times bigger then it's just you know two times bigger you have a bunch of changes in a factor of
two or three or four from each other cross staking messages is much much more powerful and that's really the word we're looking at and say hey at the
point if you know the Hub and you can live the just to this non-hub world of four or five change you have secret you have osmosis you have Juno you have
injective maybe I think that all similar market caps are already in cosmosum right and that together then that mesh
security would say okay this is actually a very a different analysis which is different than where they're coming from so but I mean I give the same thing if
you the highest security can get in the uh in the um in causes right now is 100
secured from the cosmos Hub currently yeah right yeah but and they and they just I think there's a thing of like
informal uh you know I like them I like their work IBC rust big fan you know
but it's like that they it I I think they're just very very blinkered by the focus on the Hub because they're
economic incentives are so so heavily aligned with the fact that the Hub has
to be a hub and it's like it's now at the point where people who think like
that that's fine that's just that if this is just your opinion that's cool but if it's if it's going to get in the way of
like the the development of the cause most more generally and like like [ __ ] on other ideas it's like
okay like chill out it clearly clearly isn't the the future of the cosmos clearly and self-evidently has to be
that these other ideas that are more about smaller lower market cap app
chains working in heterogeneous ways that are talking to each other by protocols but have internal
implementations that are vastly different that is the only way the cosmos makes it and if if you really
just want to only focus on the Hub that's cool but that will bring the whole ecosystem down if you win if it
becomes a question of do the innovators win or do the incumbents win then the incumbent winning would mean and the
cosmos not making it right so I think that's why I'm being a little bit of a like free son of a prick about it because the the danger of letting of let
like it's just like a minor bit of thud because they're like oh the idea isn't fully threshed fleshed out and you're like yeah fine it's kind of natural to
to their incentives are we've been working on this thing for a year and a half and we want to just be like hey come on guys yeah don't get ahead of
yourselves but it's also like very I think it can also be quite Insidious to let those kind of ideas go a little bit
unchallenged in this sort of a discussion because there's a lot of there is a lot of money and there is a
there and there is actually you know kind of the future of the ecosystem at stake in terms of like we you know
you do need to encourage people to keep an open mind about these new developments that are coming along right and and to be fair in in this document
I'll put it in the link in the show Notes too but in this document they they do recommend a switch to mesh security when it's ready like even with that 10
that that I think Jake called out on the power cap and so they're it's before
their math class at the end but before that they do say like when it's ready switch to it so and that's early right
so I I mean credit you know they recognize it yeah yeah I I don't like to
know as well I just want to say that it's fun to criticize them I think you know the more people we have looking at
this stuff the better um if I wasn't so busy like trying to like build [ __ ] I'd spend more time like
engaging you know with like responding to the analysis but I think you know
this is the kind of feedback that makes like you know like protocols like mesh security like a lot stronger and we'll
be happy to work with them as we get like closer to the final implementation the only thing I didn't like was just
like yeah I don't think they calculated the math right um but you know can be fixed can be fair
just want to say as well on the point of Ben generally making decent points in the comment he also observed that null
likely holds more uh fought than atom and that and he's really actually just
excited about the return of the castle which I can confirm although to be fair maybe that's more jabby actually jabby
was the real Javi was the real forces degenerate he nearly ruined himself on that one
um that he went deep on the castle he went deep into the castle lads
um okay it was so funny at the end of it we're like man we've been saying it's a
[ __ ] Ponzi the whole time what are you doing and he's like yeah I thought it was a Ponzi he even knew it was a
Ponzi on on the way on the way up and on the way down so
um I've got a question there from soy two Studio because he was he was talking about some um some questions that Nick
white of Celestia is a bit of a Critic and so I've asked for some more pointed questions so that we can bring them up
um so one question he's got here uh in case of an attack you would have to Fork every chain with malicious cross stake
on it to penalize them anyone want to um add some commentary to that I don't
believe it you mean this talking of attack is a successful Byzantine Fork of
a consumer chain I I think I think what he's talking about is that if a consumer chain were to you know have some sort of
wasm attack that may have been successful and altered the chain or something similar I think I think
remember correctly I I think his point was that if a provider chain is actually
malicious it will just censor any evidence from being submitted to it so
that their own validator is going to get slashed so if like I have a chain that's like you know providing security to
another chain and then you know I can just like prevent any of the evidence coming in which is a couple of things
which is one I mean yes you just chose a bad provider chain right and the whole point is that you know you should not be
relying on only one provider chain to provide your full security right you should be relying on many provider
chains to do that so that way there's not just one chain can do that the second response though is we actually so
uh thinking about this exact problem actually led us to come up with this uh new idea that's like kind of helpful for
dealing with censorship of mesh security evidence but actually just helpful for
dealing with censorship of any kind in Cosmos and so we're calling it interchange sanctions and how that works
is let's say you have a chain that's like let's say osmosis is like censoring
someone from doing from submitting certain from submitting transactions right I mean one thing is that you know
the goal of our threshold decryption and all that kind of stuff is to help solve that but on but secondarily like you
know what you could another thing you can also do is submit your transaction let's say on Juno your osmosis
transaction you submit it on Juno into the IBC Channel and what it will do is
any new IBC messages that come we today we have this like uh in IBC you either
have like unordered channels or fully ordered channels we're proposing this like model of like partially ordered
channels where you can um say like hey I want to make sure this transaction gets at the top of this IBC
queue and block any other transactions from a package from going on this channel and what you can do is basically
if you feel like you're being censored you just submit your transaction to all the counterparty chains of the censoring
chain and it basically halts all IBC communication to the censoring chain until they process the uh the
transaction that's being censored so it's kind of like interchange sanctions where all the chains are sanctioning uh
they block off IBC communication to that one chain until that chain stops
censoring basically so I think that's actually so you know the original goal here was like hey we can use this to
make sure it changes on sensor mesh security adjectives but there's other you can use this for any sort of censorship really
uh I think another point is like if if a provider chain did prove malicious like
well consumer chains have options they can just like remove remove that provider chain now that provider chain
is not getting any rewards um and then if like a provider chain was doing this a bunch of people would
probably like remove that provider chain um and so there there is like you know sort
of um some Game Theory there um but
yeah I think a lot of people that are critiquing um mesh security have like huge bags in
something else and they uh tend to you know interpret mesh Security in like
like a very broad way or like just the way that they want to interpret it like a lot of the early critiques just
assumed it was like based around validators which it's not like um and just a lot of like people teams
didn't watch all of Sonny's talk because they like missed the parts about like voting power caps and you know
um things like that and so pretty short of digestible talk I mean so I I bang on about about how the
evidence is that you can easily digest about 20 minutes of of technical
information with somebody talking at you before before you start to basically drop off your brain stress level goes up
or whatever Sonny's talk is almost exactly the right length to digest so you've really got no excuse
me I don't know how you do it you're like you lay out the most complex in like the most clear way possible I'm so
jealous I think it's that that [ __ ] Armor was so [ __ ] heavy you just need to get it done and get the [ __ ] out of
it right yeah hey Sonny I have a absolutely completely
unrelated question to anything we've been talking about um you did it you did like a presentation
I'm pretty sure uh last year sometime maybe even earlier that was talking
about the um lock-up period for tokens and why it
exists for the security do you remember what that was on so I can bloody well look it up on YouTube
um I think it might have been my Cosmo verse talk from last year when I introduced super fluid staking I'm
pretty sure and I just could not bloody well find it
yeah I think you just look up the class for verse 2021 talk on superfood staking and like I think the first half of that
is about like bonding and unbonding periods yeah okay I'll try that thanks
we should open the betting pool now you know uh customers 2021 suddenly dropped supervillains taking 2022 he drives my
security 2023 what's he gonna drop like what crazy Innovations is essentially gonna drop in 2023 because reverse which
made some sort of you know do we have a Futures Market there some protection Market on Cosmo somewhere I mean it
depends it depends what what happens on Europe's eastern border it might be it might be here's how you cook a rat on a
spit over an oil barrel in the ruins in the ruins of a major American city but uh it's like that could be the hottest
technology of 2023 uh introduce mesh to dodge
yeah yeah right yeah hope that Australia will
still be standing at least yeah have you ever read the book uh on the
beach that's uh that's literally the premise of the book the premise of the book is uh all-out nuclear war to quote Defcon
um and uh Australia untouched but obviously that nuclear winter is not
survivable so it's basically about a bunch of Australians going like oh no we are actually [ __ ] oh well
better get a point how much beer do we have left that would be my summary of the book there's like
just a lot of there's there's like a bunch of like American naval ships that are kind of caught out in the Pacific that just rock up to Australia and
everybody's just like yeah you guys really [ __ ] it anyway how long we've got left to live three months oh [ __ ] okay right I guess I'll
go walk about for a bit and then I'll have a point and take some Valium then Australia on its own doesn't have like
enough resources to like can't it is it's not self-sufficient enough so
um during the radiation level the background radiation level goes up so much that you have a like a statistic
like there's a statistical um likelihood of you developing like life-ending cancer that is like
is this in the book you're talking about yeah
building he's like dude next year nothing's gonna happen we're all gonna die
do you know um fun fact it could be I think you're wrong hey look you set your
expectations low right if suddenly if we're around for
sunny to give a talk we're gonna be like well man it wasn't mesh security but like we're all just happy to be here man
and then if and if Sonny does do an amazing Epoch changing talk we'll be like smashed it mate another another
classic there's a fun fact about Australia um with regards to like you know our
policies around uh you know getting taken over in wars and stuff and people rocking out with guns uh so in World War
II we just drew a lion like down the down the Eastern side of the country and we're like well if it all gets too much
it will just blow up all the access across the Hill from like this side to that side and we'll just call this
Australia over here just cut it in half did it just go like [ __ ] it you can have this part it was it
was the scorched Earth policy so similar to me with my lawn that's what Russia did wasn't it
basically move everything over the Euros um and which actually was a an interesting historical point is that
when the European Empires were at their height the rush now aristocracy uh felt
um like fomo I guess for not having their own exploitative Empire so they
put a post in the Ural Mountains and on one side it said Russia and on the other
side said the Russian Empire citation needed this was something this is something I read in as a seminar reading
when I was at University as a tangential thing but I'm I'm certain I didn't [ __ ] hallucinate this uh and it was
so that they could basically be like no no at this exact post at this point up to here is European Russia it's like
European Russia this is Russia Russia and then here is the Empire I mean they're the same contiguous bit of land
but believe me they're they're definitely different um so it sounds like you kind of did
that except without the post about them it was like I will let that big go that's fine we've got another one we've
got another one it's better one over here like that was last year's
okay cool um sorry this is this is the problem this is why this is why we have normally
the one hour limit on guests because after an hour our attention span wanders onto random subjects
that's the danger um meanwhile I'm Googling you're all mountains Russian post
guys I really appreciate you guys coming out and talk about this this is the first time I really understood it and I know reached out to all three of you
asking questions and things like that that's awesome thank you
that's right uh do you see this is right yeah I was going to see you guys and good to see you Jake and sunny
is pretty cool yeah yeah this is fun it's been an
unusually informative episode back to our usual [ __ ] next week
yeah I think somebody said in the chat a little bit earlier um something about Sam Bachman Freed's
haircut and the panel and I was just like we were literally talking about that before the show started whether all
uh because I think jabby just said probability for present surprises approaches one as expectations go to
zero as stress levels go to one probability of Sam bankman freed haircut also goes to one they are they're
one-to-one correlated strongly positively correlated uh the longer the bear Market goes on
the more the more curly hair you sleep that's weird not just curly more stressed haircuts
stressed beards stress beards stress haircuts um so
so how I think usurper in the private chat the private chat we're showing you
behind the curtain in the final few episodes of the show viewers don't get too too freaked out by these Revelations
professionals what do you mean the final in the final few episodes are we not doing more episodes
for new episode wants to know what is a reasonable
expectation or time frame for mesh in the wild I mean in terms of like production
like real production to me production is like high economic value secured between like big chains like osmosis and like
Juno for example like yeah it's it's hard to say but definitely like that's gonna take a lot longer
before before we deploy it in a way that people are gonna like really ape into it like it needs to be like very very
battle tested we need to like work on the ux around it we'll probably need to get like multiple Auditors or whatever
so like I I don't imagine that before like Q2 at like super early early
earliest like he ends probably likely like after that in terms of like
uh [ __ ] around and find out test in proud prototype end of the year I'm serious like we're probably gonna
[ __ ] make this happen oh no yeah don't yeah
[ __ ] around find out in 2022 wow yeah I think I think uh I think it'd be really fun to just like do it between like Juno
and Wawa or whatever and we'll like even like cap the amount of rewards so it's like not like
well we'll just like put some hard caps in there to like keep people safe but I think you fundamentally just like have
to like you know test these things out with like small amounts of like real money otherwise people don't have
incentive to like try and break them so um I think yeah uh I look for that by
like the end of the year some wedge somewhere Jake's lawyer is having a [ __ ] Hemorrhage like at this exact
moment it's gonna be like does anybody seen the
um the GameStop documentary it'd be like that you know where they have to set it hearing and they just like replay this
clip where Jake's like at some point you've just got to play around with some amount of real money and see what happens they just like the screen terms
of service just look where you know right use at your own risk great he really said that he said that he said
that meanwhile Jake's likes Jake's there on the call with like the poster of the cat saying hang in there in the background
like I like the mesh I like them one thing people maybe don't know is I
think we rewrote like probably like 60 to 70 of like super fluid staking like
two weeks before the launch of it uh so uh the trick to getting things shipped
is you bet body parts last time which was my uh maybe this time we do things you bet body parts
yeah the other thing is getting a few amazing devs together locked in a room
with a bunch of caffeine yeah I think that uh that speed things up a lot no meetings no internet uh no calls no
social media no messaging just GitHub and code good
yeah back atopia test net huh I mean dangerously close to actual
validated chat but oh it could Topia like I'm a Public Health Net there's a
yeah there's a there's a testnet we're on the short list I think I think there's some people mad compound
staking and I think we're out of the valve set already because we're just like it's a test that we're not gonna [ __ ] run an awkward compound stick
test Nets it's just like oh my gosh guys it's unbelievable idiotic I was like trying to be [ __ ] nuts I'm not
[ __ ] doing and then yeah we're out of there I'm just like okay well if you have challenges to do whatever [ __ ]
text me or something I don't know it seems like really it they've got a good product but it does seem quite
disorganized would be my not my lukewarm take on how they're doing some some
people were going [ __ ] Ham on uh on defund sets as well so we just like did
an upgrade and turned like um inflation off I think Old Testaments
should have inflation there's no reason to have it on yeah so on on Juno I'm actually trying to out
compound the the prolific compounder and there's still that one [ __ ] guy I've
I've not checked in a little while yeah but I'm I'm Auto compounding to the entire rest of the [ __ ] set from the
bottom up so like I've almost got him done it's uh should adopt like uh on our spell says
we only give out the staking rewards once per day because otherwise you know it's this game of having to like Auto
compound every block when if you just do it once per day then you can just give it to like restake or something and they
can just Auto you know artsy and they just Auto compound everyone's once per day which is like okay I think the right
way to do it he pops all around foreign