05:03
ᴅɪꜱᴛʙɪᴛ ⛓️💵
Where can I read more about this and the reasons for it?
05:04
Zack
It was something about state actors controlling the majority of exit nodes, right?
05:04
ᴅɪꜱᴛʙɪᴛ ⛓️💵
Ah the Sybil attack vector?
05:06
Ilmu Somebody
Exit nodes owned by states, correlation using timing of packages is also used by global passive adversaries
05:06
There's also fingerprinting based on behaviors/configurations
05:07
ᴅɪꜱᴛʙɪᴛ ⛓️💵
Yeah that is a difficult problem I think, although if the effects of leaking packet routing data are able to be detected by others, then maybe a wot could help here. Otherwise a wot is useless as there is no way for people's reputation to be impacted by selling their node's network traffic if it can't be proven or shown to be probable.
05:08
It does seem to offer significant enough benefits to dnms though for many to exist for several years and for them to all have unanimously decided to use Tor over clearnet
05:08
So it must have some value?
05:08
Ilmu Somebody
But what is the point of discussing onion / mixnet routing? For physical goods that clearly doesn't work the same way
05:09
Sure. Even if NSA knows some smugglers they don't have the Overton window to use the information
05:09
So they focus on bigger fisj
05:10
ᴅɪꜱᴛʙɪᴛ ⛓️💵
Why do you think this is though? Initially this was my assumption but after reevaluating my logic here I have changed my mind
05:10
Ilmu Somebody
Any interaction with physical world will require some trust to be in place
05:10
ᴅɪꜱᴛʙɪᴛ ⛓️💵
As in trusting someone to hold your packages for you?
05:11
I think that can mainly be aleviated using escrows and bonds
05:11
As it is pretty easily verifiable whether someone stole your parcel
05:11
Ilmu Somebody
Because as an intermediary of a package in a mixnet the package is in an unbreakable infinitely thin container that I can take off only to have another visually distinct infinitely thin unbreakable barrier with the next sender address
05:12
In the physical world other laws apply
05:12
Ehhh ok so trust can be used to make trusted interactions work
05:12
Z
Anyone interested in helping me organising a really secure crypto chat? Trying to organise an opposition to the current crazy going on in Italy
05:12
ᴅɪꜱᴛʙɪᴛ ⛓️💵
Well more to the point reputation. 👍
05:13
Z
I’d need a really secure way of assigning every one citizen one vote
05:13
Ilmu Somebody
Sure. Do you have people to communicate with?
05:13
ᴅɪꜱᴛʙɪᴛ ⛓️💵
Could you elaborate on what you mean specifically by this, and how you think it is specifically different in terms of vulnerability to certain attacks?
05:13
Z
(Sorry to derail your conversation, if you prefer we can talk about it some other time)
05:14
Zack
you can know that, but you can't prove it to a third party, so it isn't enforceable. The oracle can't tell who lost the package.
05:14
Ilmu Somebody
If you read how the nym mixnet does routing and why you can see some attacks they are trying to solve for
05:14
In the physical world I can weigh the box, break the box, etc.
05:15
Zack
I wrote most of a chat server that uses off the record encryption, and it also lets you send the message to a group chat, so more than just one other person.
05:15
Ilmu Somebody
You'd need pretty advanced infra in place to replicate the coordination required to "mix packages"
05:16
Ilmu Somebody
And usually the things you are sending are either all the same (roughly; drugs, guns, slaves) or not in need of such measures
05:16
Z
Very interesting! I’ll let you guys dig into what’s going on right now and get back to you if that’s ok
05:16
(Really really noob in programming)
05:16
Barely a script kiddie, and i’m old(ish)
05:17
Ilmu Somebody
Where in Italy are you?
05:17
I live in Slovenia we could meet up if you want to chat
05:18
I'm very interested in helping the underground in Italy organize
05:18
ᴅɪꜱᴛʙɪᴛ ⛓️💵
Well you can I think use heuristics like if a package is lost on a route, decrease your subjective/relative reputation of each hop on the route, and then if a node is consistently malicious in many routes then they will eventually lose their entire reputation. At least that is a worst case scenario but there are probably some other heuristics you could use as well to make it a bit more accurate
05:18
Ilmu Somebody
You are slowly getting closer to datalisp
05:22
ᴅɪꜱᴛʙɪᴛ ⛓️💵
If a parcel is lost in between node and a node b, then one of them is responsible for its loss so both can be punished a bit... Maybe some mechanism could be designed that allows for nodes to specify certain nodes who they don't want to be after or before in a route, due to them not trusting each other in the wot,
05:23
But how to enforce that is difficult
05:24
Zack
The attacker is damaging others equal to it's own damage. it is a griefing vulnerability.
05:25
ᴅɪꜱᴛʙɪᴛ ⛓️💵
But due to the attacker being in many different routes over time with different nodes, it isn't a sustainable griefing vulnerability
05:26
It is only a temporary griefing vulnerability
05:26
For as long as people still trust the attacker to route for them
05:26
Z
Thank you! I move between northen and central italy, will be in serbia in october if that is more comfortable
05:27
Ilmu Somebody
Cool you can pm me and we can coordinate
05:27
Z
Gladly! Speak soon
05:29
Zack
yeah, that makes sense.
I never realized how symmetrical the data privacy and physical privacy situations are.
05:30
Until I thought a bit more about it yesterday afternoon
05:31
It was actually due to something
@jmharvey said
05:31
Which I tried to write a response to explaining why I thought Tor hidden services were fundamentally harder to restrict and regulate than physical drug delivery
05:31
But then I realized I actually didn't have any case
05:37
Of course sender privacy has always been extremely easy to achieve in the sending of physical packages
05:37
What had always been hard though is recipient privacy
05:38
Which I think onion shipping solves
05:40
Ilmu Somebody
Yeah hidden services are /hard/
05:41
Have always relied heavily on trust (codeword into back room etc.)
05:43
ᴅɪꜱᴛʙɪᴛ ⛓️💵
By this are you referring to the issue of nodes undetectably leaking routing data to adversries?
05:44
Ilmu Somebody
No I mean literally
05:44
Revolutionaries in the middle ages or something
05:47
ᴅɪꜱᴛʙɪᴛ ⛓️💵
Wouldn't it have been harder to run hidden services back then due to not having access to modern cryptography, digital money and digital coordination mediums?
05:48
Which all make routing a package between many psudonymous actors much easier
05:48
Ilmu Somebody
Sure .. I'm just saying that the way it's been done is usually based on trust or security by obscurity (timing sensitive)
05:52
ᴅɪꜱᴛʙɪᴛ ⛓️💵
Ah right yeah it seems that trust is definitely a necessary element in such networks
05:53
It has always been the case in tor, but it is just now even more obvious than ever. Unfortunately though most of that trust in tor's case is vested in the foundation who controls who is and isn't able to run a node
05:55
But Sybil attacks are always possible as an attack against obfuscation technology
05:56
But some are a lot easier to carry out such an attack against than others
06:48
Josh
Hehe, what did I say?
07:36
ᴅɪꜱᴛʙɪᴛ ⛓️💵
This :)
19:33
Zack
This is a place to talk about amoveo and related game theory.
Miles Anthony invited Miles Anthony
17:30
Zack
For any Dapp, we have the option to either:
1) make the contract state be a part of the blockchain's consensus state.
2) have some deterministic way to calculate the contract state by looking at the blocks.
Why would we use (2)?
(1) puts more cost on the full nodes.
With (2) you can change the rules for calculating the smart contract state, without needing to update the full node. As long as all the users of the smart contract can agree on how it works.
Why would we use (1)?
With (1) we can use the currency that was produced from PoW mining as collateral to enforce contracts.
Im thinking of something like ERC-20 tokens on Ethereum, or NFTs. They are currently a (1), but maybe it makes more sense for them to be a (2)?
17:33
To use a (2) of erc-20, you would need to sync a second light node, for that contract.
So if you wanted to use N currencies, you are only syncing N light nodes.
17:48
(2) is how all contracts would work in a optimistic rollup blockchain.
17:49
Making an erc20 on ethereum is so expensive. Adding this feature as a type (2) on amoveo is pretty easy, and the amoveo codebase probably doesnt need to change at all.
17:56
ᴅɪꜱᴛʙɪᴛ ⛓️💵
Wouldn't this be somewhat different in that the mainchain eventually achieves consensus on what the state is of the OR (as in, when people withdraw)
17:57
SLP tokens/counterparty/omniledger all seem to also fit into (2) right?
18:02
Wrt the distinction between 1 and 2, so I am clear on it, is it that miners and full nodes don't enforce correct state transitions in (2) but do in (1)?
18:26
Zack
The mainchain eventually achieves consensus in (2), yes.
18:26
No. They enforce it in both.
18:27
The distinction is that (2) is not in the consensus state. It is something you can calculate by applying some rules and scanning all the blocks.
18:28
So you would need a second light node to be able to see into (2)'s state.
The light node for the main chain wouldnt work.
18:30
Counterparty works this way on bitcoin, yes.
18:39
ᴅɪꜱᴛʙɪᴛ ⛓️💵
Does that mean that smart contracts that influence the state of the native asset can also interact with the erc20?
18:39
Or are they necessarily causally disconnected, except for via the use of oracles?
18:43
Wouldn't it need to be a second full node, as even if a maj of miners are honest, since they aren't enforcing the erc20 protocol due to not having the state, invalid erc20 txs can exist onchain?
18:43
So you would need to see all txs in order to determine the current erc20 state?
18:43
So as to ensure you haven't missed any that could be of relevance (for example by being earlier than a conflicting tx which you thought was valid)
18:48
Zack
Everyone can see the type (1) contracts.
But type (2) contracts can only be seen by themselves.
18:49
The layer 2 full node needs to see all the state.
The layer 2 light node can trust layer 2 full nodes, but if there is a disagreement, it downloads only the block that they interpreted differently. So it only needs to download one block for every dishonest layer 2 full node in the list of who it downloads from.
18:54
With this strategy, spending a new currency or nft would cost as much as sending veo now.
But the big advantage is that creating a new nft or currency could be free.
19:04
ᴅɪꜱᴛʙɪᴛ ⛓️💵
Nvm I am actually not totally sure about that (wrt my deleted message)
19:04
Because of the ability of previously unavailabile data to become available
19:05
Free as in it doesn't require the creation of a new chain specifically for it?
20:09
Zack
You do have to make a new chain. But it is free. There is no tx fee.
20:22
ᴅɪꜱᴛʙɪᴛ ⛓️💵
Hmm ok I think I am misunderstanding something then. How is data able to be added to the chain without a fee needing to be paid?
20:23
Wouldn't a tx in a light client validated erc20 chain need to stored on the amoveo blockchain in order to acquire an objective ordering between itself and all other erc20 txs?
20:24
Zack
Creating a new chain doesnt add data to any existing chain.
20:24
ᴅɪꜱᴛʙɪᴛ ⛓️💵
I wasn't awere that this proposal was for creating an entirely new chain
20:25
Zack
It costs thousands of dollars to make a new erc20 on ethereum
00:05
Zack
If we wanted global harberger land taxes, what would the database look like for assigning land ownership?
I think there is a point-line projective duality we can take advantage of. Every point on the sphere has an opposite point on the other side, and between the 2 opposite points is a line. like how the equator is a line between the north and south poles.
So if we have some coordinates to talk about a point on the sphere, those same coordinates can specify the line.
Instead of worrying about the points at the corners of a particular property, we should pay more attention to the lines that connect those points.
Every line is a great-circle that cuts the world in half.
These great circles are a good basis for the database, because as we walk down the merkle tree, each step can be a great circle that cuts the world in half, so that we are certain each of the 2 branches do not overlap at all.
Each location on the sphere can only be in 1 location in the merkle tree.
So what I am trying to solve for now is what format should we use to refer to a particular great circle? I am pretty sure it is a 2-partnumber, but maybe both parts can be fractions, so it is better to make it a 3 part number where the 3rd part contains the denomenator part.
We want to choose a format that makes it very easy to calculate the point where 2 great circles meet, and to calculate the great circle formed by 2 points.
00:11
The great circles on a sphere, they can be mapped to any planes that pass through the center sphere.
I think if we look at where these planes intersect the plane that is tangent to the sphere, and touches the north pole, then we can convert the complicated sphere geometry back to normal rational geometry, and maybe this is how we can get the best coordinates.
00:13
kind of like how the klein model turns hyperbolic geometry back to rational gemoetry, I think this projection is turning spherical geometry back to rational geometry.
00:42
Zack
so, the equator gets mapped to the north and south pole, and is (0,0,1)
great circles that pass through the north and south poles, on lines of latitude, they coorespond to points on the equator, and in the projective plane they may to points that are very far from the origin. something like (9999999,0,1) or (0,99999999,1).
points at 45 degree latitude (so, in the northern hemisphere) correspond to great circles that pass through 45 degrees at their most northern location, and 135 degrees at most southern. So, locations like (1, 0, 1) or (0, 1, 1) or (-1, 0, 1) or (sqrt(2), -sqrt(2), 1)
In this format we can only refer to pairs of anti-podal points, so the points at 45 degrees latitude are the same as points at 135 degrees latitude.
00:49
how many layers does the database need.
earth's radius is 6371 kilometers.
the surface area is 4*pi*r*r = 5.1*10^8 square kilometers. or 5.1*10^14 square meters.
each great circle is a binary split. so to divide down to 1 square meter, we need 48.8 steps of merkle proof.
Each step of the proof is 256*2 bytes of hash of children, and maybe 16 bytes to specify the great circle. = 528 bytes.
so a proof is around 50*528 bytes, about 25 kilobytes.
And that is the worst case scenario where we divided up the entire planet into square meter chunks. Assuming some people want to own more than 1 square meter, we can have shorter proofs.
00:58
The biggest limitation we still face for this harberger land plan, is how to do the combinatorial markets for changing ownership.
It seems like we need some way for people to store their bids, and some way for people to propose combinations of bids that reorganizes ownership of land in a way and results in higher total taxes being paid.
01:00
It probably needs some way to cancel bids too. and bids should probably have a time limit until they expire.
01:03
We probably need some system to pay people to rebalance the merkle tree as well.
01:08
Maybe it makes sense to cut the world into chunks, and then have different sidechains for each harberger chunk.
the advantage would be that sidechain full nodes are cheaper. the disadvantage is that the combinatorial markets wouldn't work at the borders between the sidechain regions.
01:19
So, what does a bid look like?
I guess a list of points or lines that encircle the area that you want to own. Points and lines are duals, so they take the same amount of space to store.
How would we store bids in a merkle tree?
What do we do about expired bids?
01:23
We probably need a bid where you are also willing to buy any subset of the space you are bidding for, at the same price per square meter.
01:32
Maybe we need a new tx type in amoveo that is just a comment. So you can write info that is used in the layer 2 harbinger system.
13:46
Ilmu Somebody
This will never work, it's a regression from what we have now
13:46
Please read seeing like a state
13:48
Land ownership has been oversimplified already and the last thing we need is another model that is too simple to handle the complexities of what the reality on the ground is like
16:02
Šea
hey, how's development going
16:02
did Amoveo team expand over the years
17:15
Zack
It seems like the project that gets harbinger to work will win, because they will have so much liquidity to work with.
Pow mining can increase the money supply, but the only sustainable way to have a big supply is if there is something valuable backing the coins.
17:16
It is going ok. We are mostly focused on the dex for now.
It isnt clear how many people are involved, since almost everyone is anonymous.
18:50
Zack
The rise of harbinger land registries might end up being similar to the enclosure movements throughout history.
The new system of resource management is far more efficient than the old, so there is a big economic incentive to convert land and resources from the old system to the new.
Anyone who is able to capture and convert land can profit considerably.
I wonder how things like oil fields and hydroelectric dams and canals will make the transition.
I imagine that in most cases, the existing governments will actively participate in selling out public resources, so they can take the windfall profits as a corrupt bonus.
But inevitably, some regimes will resist the transition. Creating a large prize for whatever violent group manages to wrestle control of the resources away to be sold off to the harbinger system.
23:31
Zack
How about if there was a way to settle the contract, so instead of distributing the money to subcurrency holders, the contract just says an address, and all the money goes to that address.
This is how bitcoin contracts work.
For bitcoin spend transactions, the transaction signature is part of the input data for running the smart contract.
So we could reproduce all the bitcoin contracts.
Multi-sig for example.
Deleted invited Deleted Account
04:10
Zack
hi. it looks like that private key doesn't contain funds.
Maybe PM me the private key directly in the future instead of putting it in public chat.
04:10
Deleted Account
It does
Akane Yokoo invited Akane Yokoo
12:54
Akane Yokoo
Greetings admins, i represent dex trade exchange, could I speak someone who is responsible for listing of this project?
Thank you
15:17
Ilmu Somebody
What kind of shitty DEX needs a centralized authority to approve listings 🤔🥱
Deleted invited Deleted Account
08:26
Zack
Our centralized institutions are all at risk of assassination contracts.
If we quickly upgrade everything to be decentralized, and we use futarchy to make important decisions, then we can patch this vulnerability before anyone gets hurt.
08:38
ᴅɪꜱᴛʙɪᴛ ⛓️💵
If we had hierarchical decentralized instructions though they are subject to ACs to the extent that the leaders aren't able to maintain perfect opsec/pseudonymity
08:43
Zack
I wonder how much the anonymous hierarchy strategy will be pursued.
It isnt compatible with democracy, and it seems like the kind of thing that turns into a dictatorship quickly.
08:44
ᴅɪꜱᴛʙɪᴛ ⛓️💵
I was thinking even in terms of company structures or private security agencies
08:45
It seems the market has converged on a fairly hierarchical structre for companies, so making them more leaderless would probably decrease their efficiency quite a bit
08:46
Zack
If the person making decisions for a company is anonymous, how can there be consequences for making decisions that harm shareholders for personal benefit?
08:48
ᴅɪꜱᴛʙɪᴛ ⛓️💵
Reputation+collateral/bonding+loss of job
I don't necessarily believe that it is a solved problem though, despite many daos beginning to exist. I don't think they are really DAOs though in any useful sense as most don't really solve this issue in a decentralized manner
Deleted invited Deleted Account
21:52
Deleted Account
Hey guys, my friend gave me a private key of $4000 in a wallet but I don't know how to access it, this is the private key:
77da1482ecf0306db09f5ae0dda3742eed7320f1132cb9d84853b29ac4777a73
Please can someone explain to me how to use it
Dereek69 invited Dereek69
23:02
Zack
Why is she asking about a completely different key this time?
23:03
Is this some kind of scam?
23:05
Someone went through the effort to make something that is formatted like a private key.
07:16
It is kind of important for us to find out if this can work as a counterparty style sidechain, or if it would need to be added to amoveo as a hard update.
07:17
The part that I am not sure of is when people make bids to buy regions of land.
Doesn't a bid require locking up some veo that could potentially be used to pay for the land?
07:18
Especially since we require combinatorial bids.
Because sometimes you need to own a particular combination of land to be able to execute your business plan. so owning 1/2 of your region is worth far less to you than 1/2 the value of owning all of it.
07:21
It uses 12 bytes to record a great circle.
07:24
if the merkle proof had 35 steps, we could cut the world into acre sized patches.
07:26
Maybe the bid is somehow a swap offer, and making a combination of bids is like, making a multi-tx with all the swap offers in it.
08:05
Zack
I wonder if the best plan is if the harberger currency is different from veo.
So veo is getting value from pow, and from being the only currency to operate the oracle.
And the harberger system gets its own value from the land taxes, and being a sidechain of amoveo, it has access to all the oracle data.
With that design we might move the smart contract system and amm and swap offer tools all to the harberger side, and have veo be just for oracles. We cant get oracles off the root chain because of how we are mixing the oracle mechanism with the consensus mechanism.
Having the second currency would be optimistic rollup style scaling.
If we can use real veo transactions to do the land purchases, that would be more like counterparty scaling.
08:12
It would be really nice if the land and pow were hooked up to the same currency. Because then the land is giving a consistent source of value for a calculatable market cap. And the pow is for smaller optimizations, to rebalance the market cap if the demand for the currency suddenly grew.
08:17
Oh, there is also that problem that the oracle reports can need to grow as big as all the money being bet.
08:17
So we need the oracle currency to be the same thing as the harberger currency.
08:18
It would be really nice if amoveo full nodes didnt need to maintain a database of land registries.
08:18
Oh, there is a way simpler way to do this.
The full node can choose to be a light node for certain parts of the consensus space.
08:19
So we can just add it all to the full node, and set up a configuration option for full nodes that dont want to maintain the extra database.
08:27
Buying land in the harberger system is a bet that this registry will be used to record ownership of that land in the future.
08:50
Zack
Land ownership records are currently a terrible system. Ive seen it in different countries, it is awful everywhere.
Transfers are complicated and expensive.
Databases have mistakes. Land titles can overlap.
Too many people have access to update the database, it isn't secure.
The majority of people are locked out of official legal land registries, and instead depend on local informal systems.
Sometimes the local informal systems are better enforced than the national system.
You need to pay professionals a significant portion of the cost of the land, just to help the transfer happen correctly.
There is no reasonable way to buy up adjacent land for a big project.
And because of all this, it is very difficult to invest in land, and a lot of land is being under-utilized.
I think the governments wont oppose harberger registries either, because it will make taxation a lot easier when everyone is self-reporting the true value of their real estate.
08:54
Eric Arsenault
it seems like a very complicated problem to solve
08:55
I haven't been following this discussion much, but from I remember this was a way to grow the money supply?
08:56
Zack
It looks like the market cap of VEO would tend to be about the same as all the land registered in it
08:56
Eric Arsenault
I feel like doing that via a lending of crypto (collateralized loans) would be a way easier start since people are already doing that
08:56
Zack
So it would give a lot of liquidity to enforce contracts
08:56
Eric Arsenault
yeah, I agree... it's a nice pipe dream, but feels super far out imo
08:58
it feels like there would be a massive barrier to entry
08:58
Zack
There are some papers from google search showing other people working on land registries
08:58
Eric Arsenault
yeah
08:58
Zack
It is a big barrier to entry getting Canada to enforce based on a blockchain land registry
08:59
Eric Arsenault
people have talked about it for years
08:59
Zack
but how about if we started with some poor community in India that doesn't have access to the formal state tools, and they already have lots of informal records of who can use which land
08:59
Eric Arsenault
it's not impossible, but just feels super hard. who's going to go to india and do that
09:00
Zack
they can just copy their informal records to the harberger database, and the value of their land could immediately increase 100x by being tradeable
09:00
Eric Arsenault
I mean I'm all for this sort of system, I just think it's hard
09:01
and amoveo doesn't have a strong track record of doing anything well other than innovating on blockchain stuff. I don't think we would be very good at trying to take a real world idea like that to market
09:02
You could take a lean startup approach: create a landing page on wix or something, talk to people, see if they want to do it
09:02
Zack
I feel like the 2 big limiting factors that stop other land registry projects are:
1) getting the database right. Merklizing chunks of land is a little complicated.
2) getting the game theory right. The kinds of people who care about the unbanked tend to not be the kind of people who understand economics.
09:02
Eric Arsenault
you are right
09:03
but I still think taking this sort of product to market is really hard. you need lots of funding and a team, and partnerships and very clear trustworthy entity to at least represent the concept
09:03
I mean, maybe we could partner with an entity that uses our tech
09:04
Zack
If Amoveo is going to be the biggest derivatives tool, we need a sustainable way to have liquidity for the derivatives.
I don't want to just make a big cryptocurrency.
I want to make the big global one. A world reserve currency.
09:04
Eric Arsenault
but still, feels really hard lol
09:09
it just feels like we need a bit of traction on something
09:13
Z
DEX first?
Even if you find a way of parcelling land the only real client might be some UN program
Even if X region would adopt it the land transfers would have still to be recorded in the central government database
09:14
Bets or wagers accessible to everyone would be nice also
09:15
Zack
Yes, the DEX is first.
The harberger plan is still in very early speculative stages.
There are already parallel land registries, and in many parts of the world the official central government one is the less important one.
09:15
It is really a hassle trying to buy land, and you want to own it on both registries, so it is like you are doing 2 purchases at once.
09:18
we do already have some nice betting tools. Would you like help testing them out?
09:24
oh, it would be nice if there was a way to post a bet offer for a contract that doesn't exist yet. that should be pretty quick too.
09:28
I study projective geometry sometimes, and i wrote a merkle tree database system before.
I think im in a good position to make this database.
And im feeling confident that harberger with combinatorial bids is the system that would win.
It also seems like there is a good chance that this will be 10x better than existing registries, and so spread virally. Like the enclosure movement.
09:33
Maybe a better parallel to draw was when Alexander was building his Greek empire.
How he kept capturing golden art, and converting it into currency which was much higher value because it was more liquid and cheaper to store, and using this currency to fund his conquests.
Similarly, when land gets put into the harberger registry, is is becoming more secure, more liquid, and much more valuable. So it could cause a chain reaction where people are getting control of land, putting it into the registry, and then using the profits to get control of more land.
09:41
I could buy a bunch of land somewhere and then try to re-sell it in a blockchain registry
Deleted invited Deleted Account
Deleted invited Deleted Account
16:44
Ilmu Somebody
The issue in the first place is that land registries and cadastral maps are creating a problem rather than solving one.
16:47
Zack
it seems like land is used more efficiently when the same person who makes decisions about how to use it is the person who profits from the production of it.
It creates good incentives.
Systems like this live in an evolutionary environment. If one system allows land to produce more value, then the culture that uses the more efficient system ends up overtaking the cultures that use less efficient systems.
18:04
mr x
yes... little js api from light node for creating offers, accepting offers, fetching offers by category...
18:44
Ilmu Somebody
Problem with evolutionary perspective is that it's not possible to factor everything into independent dimensions. Violence has been used to conquer land for a long time and the use of land has rarely been efficient (scorched earth is a very effective strategy for draining enemy resources for example). When land use is very efficient it creates a lot of wealth which often leads to invasions and pillaging.
18:47
The argument that monopoly resolves tragedy of the commons is not as strong as economists would like to think. Elinor ostrom got the Nobel prize in economics for observing that "a system that works in practice can work in theory". I would again recommend you to read seeing like a state.
18:48
Zack
You can make tx fees cheaper by lowering security and increasing the rate at which attacks occur.
When you calculate the efficiency of a resource, in a probabilistic setting, it is like this: sum_over_all_cases(probability_of_this_case * profitability_of_this_case).
18:48
yeah, it looks like harberger is more efficient than monopoly.
18:49
Ilmu Somebody
I mean monopoly as in ownership of land
18:49
Zack
harberger is a kind of continuous auction that you cannot opt out of.
So it enables one person to have the benefit of monopolistic usage of the land, but without imposing any cost of monopoly on society, because at any time, anyone else can use the harberger mechanism to buy that land.
18:49
Ilmu Somebody
".. the same person who makes decisions about how to use it is the person who profits from the production of it..."
Arlette Vån invited Arlette Vån
Valentyna Kamdr invited Valentyna Kamdr
Deleted invited Deleted Account
12:39
Eric Arsenault
@zack_amoveo do you or anyone know if there are any AMMs that have combined with a traditional orderbook in the crypto space?
17:56
Zack
Amm need to be on-chain, because the liquidity locked up inside them.
Order books are best kept off-chain, until an order is matched. That way unmatched orders are free.
They all work in combination together.
Marijn [REDACTED] invited Marijn [REDACTED]
Deleted invited Deleted Account
23:15
Ben
just move on. nodbody will pay you.
Deleted invited Deleted Account
22:46
BitGuru
Hello
I have a deep knowledge of blockchain technologies, specifically in DEFI, DEX, NFT projects so far. Build Smart Contracts using Solidity and deploy them to Ethereum and BSC using HardHat and Truffle, and interact web3.js/ethers.js, React.Js, Vue Js with Smart Contracts and more with industry basis. I’m full-time available and able to commit the work immediately.
Do you have any open positions that I can help you in?
Thanks!
Deleted invited Deleted Account
Deleted invited Deleted Account
Deleted invited Deleted Account
Scarlett Albia invited Scarlett Albia
18:13
Z
Wouldn’t a professional promoter avoid to repeat the word promote three times in fifteen words?
03:47
Instinct
That’s part of his unique style
Deleted invited Deleted Account
Deleted invited Deleted Account
13:07
Eric Arsenault
How are things going w cross chain
@zack_amoveo ?
13:07
I guess it’s a big update
15:03
Zack
It is coming. I am abstracting some patterns from the first 2 crosschain versions. To make it shorter and easier to understand and maintain.
I got a little distracted looking at spherical trigonometry for the harberger land tax system. I hadnt realized how beautiful this part of math is.
Now ive got a plan on how to represent land plots, how to convert from gps coordinates to the internal numbers, and back. How to calculate area of plots, distances between locations, and the compass direction to walk to get from a starting location to an end point.
The rational projective system is much more beautiful than the irrational system based on angles they teach in usa schools. Deterministic too, so we can put parts of it into the blockchain as needed.
Deleted invited Deleted Account
20:56
Sy
Imho alot (all?) of the american measurement systems are "not that logical" and since everything else is based on it this doesn't suprise me at all...
21:02
Zack
People say that babylonians didn't understand Pythagorean theorem, because we only found tables of pythagorean trios, not the formula expressing the real relationship between the sides of right triangles.
But if you look at modern trigonometry, functions like "sine" "cosine" and "tan" are just looking up values from tables.
An archeologist looking at our culture would say that we don't understand the real relationships between the angles and sides of triangles.
Deleted invited Deleted Account
Valentina Lautaro invited Valentina Lautaro
Benjamin Omar invited Benjamin Omar
धनवन्तरि कुमार invited धनवन्तरि कुमार
Deleted invited Deleted Account
Oyster Cult invited Oyster Cult
MrWerber invited MrWerber
🔥 П0UСK_PАPЫ_NА_VE4ЕP 🔥 💋 scылкa_v_пpofiле 💋 invited 🔥 П0UСK_PАPЫ_NА_VE4ЕP 🔥 💋 scылкa_v_пpofiле 💋
Deleted invited Deleted Account
Johannes Jger invited Johannes Jger
23:42
Zack
I got the key geometry result.
Given a list of great circles on the earth, like you would have while verifying a merkle proof of the land registry, it can calculate the gps coordinates of the corners of that land, and the area.
02:49
Ben
did you already finish the DEX or did you put it aside? Wasn't the dex the MVP?
02:50
Zack
buying veo and selling veo work.
trading 2 currencies on other blockchains is in progress.
09:00
Zack
the case where there are extra unnecessary lines is tough. Im looking for the algorithm to identify which lines are not borders of the enclosed region.
21:10
Zack
I can now ignore lines that don't touch the borders of the region.
but if a line touches a corner of the region, it breaks.
01:36
Zack
now I can handle if a line-constraint goes through a corner.
But if you have contradictory line-constraints, it still doesn't report that your boundaries contain nothing.
20:27
Zack
ive got the contradictory constraints.
ive got lines that touch corners.
but there is a problem now if a constraint line is centered in the southern hemisphere.
21:06
Ben
don't you think that diverting in a completet diffrent area is a bit early before the DEX is completed? The Dex could bring actualy the first time a real userbase to amoveo and lead to organic growth.
21:06
Ilmu Somebody
🙈 glhf
21:06
Zack
yes, the dex is important for onboarding
23:54
Z
Investors/adopters love road maps, especially when they are regularly updated and achieved...
a nice DEX with a simple and easy to use GUI like uniswap/1inch/etc could be a good for the coin
03:59
Zack
I think i solved all the geometry we need for this anyway.
16:42
Rom
Hi, when you start the UI design ?
18:21
Zack
Ui design for what?
Amoveos light node us live. You can try it out now.
18:23
Or do you mean the harberger tool?
That is still more in the idea stage. It needs some new databases and transaction types first before we can use it. And we probably want it usable before we make any interface.
Палестина Касинева invited Палестина Касинева
21:26
Rom
I made myself misunderstood, it is planned to make a fun interface usable by all users, even beginners.
and incidentally with a real web address and not your IP…
00:38
Cream E. Biggums
This coin is a joke
08:50
Zack
I think my legal risk is less if i dont have a website.
Since everything right now is decentralized, and anyone can spin up a node.
08:50
We do want it to be easier for beginners.
10:07
Eric Arsenault
😂 yuuup
10:08
Hopefully soon we have a dex that is usable
11:59
x
if you want beginers, it doesn't need a website or even ip address, just turn web app to a desktop app, that can be downloaded and used directly.
12:00
Zack
Maybe we should experiment with a desktop app for one of the applications to see if it gets more usage.
12:44
Eric Arsenault
Ummm
12:45
We just need to get the dex working
12:45
Zack
right. we could do a desktop rewrite of the dex as well
12:45
Eric Arsenault
Desktop apps are a ton more friction to get onboarded
12:45
javascript is nice because it works on phones too
Deleted invited Deleted Account
20:59
Ilmu Somebody
this is why telegram should support replying to more than one message ... why does everyone hate lattices...
20:59
Zack
Discord is nice how you can put a little image on a comment
21:01
Ilmu Somebody
Discord doesn't allow third party clients.
21:01
Zack
does telegram?
21:02
Ilmu Somebody
yes, it actively supports it, they maintain a great library for integrating telegram in different places, telega.el is a pretty nifty emacs integration for example.
21:02
Zack
wow, I didn't know that
23:33
Zack
I wonder how high harberger taxes should be. It seems like making them very high would make it easier to get access to land quickly to experiment with new businesses.
but having them lower could make it easier to absorb new land into the system, because the previous owner is losing less by registering.
Starting from the tax free case.
The interest rate is I.
land produces F value per year.
So then, the land is worth price P=F/I. Because people will keep buying up the price of land until the profit from holding that land is the same as the background interest rate.
What if we charge a tax of T based on the value of the land?
So people will keep buying up the price P of the land until: P = (F - (T*P))
/I-> PI + (T*P - F) = 0
->P(I+T) - F = 0
->P = F/(I+T)
So if the land in total is worth $12k. and the tax rate is 1/2 of the interest rate, then the price of the land would drop to around $8k.
if the tax rate is double the interest rate, then the price of the land would drop to around $4k.
The land doesn't drop to zero value no matter how high the tax goes.
So where did the extra
(F/I) - (F / (I+T)) =
(F(I+T) - FI)/(I(I+T)) =
F*T/(I*(I+T))
of value go?
Assuming this money was burned, then the value became part of the market cap of the currency that is used to pay harberger taxes.
00:53
Z
Discord if managed properly can be more useful to onboard new entries to the project imho
00:58
Zack
If the tax is higher, there is less incentive to improve your land. Because a larger fraction of the improvement's value goes to the cryptocurrency.
So lets say amoveo owned 90% of the value of a piece of land that the owner wants to improve.
Doesnt that mean amoveo should be willing to cover 80% of the cost of that improvement? Since the net result is increasing the market cap and value of the token.
So then, would people use futarchy to show that an improvement is good, and then get free money from amoveo to help cover the cost of the improvement?
01:03
Imagine a world where you could start your own physical location business by only paying 10% the value of the building+land to buy it, and 20% the cost of renovations.
Instead of paying rent to a landlord or mortgage payments to a bank, there is the tax to amoveo.
01:03
I guess we dont want the tax too high, because it removes he incentive to not destroy the land.
02:06
Jeans
It's awesome, but hopefully soon we have a dex that is usable
03:48
Toby Ganger
just four more years and then we’ll be really close
04:07
Z
Dont let your great ideas get in the way of developing what you have on hand.... host it ninja on the darknet if you really don’t want to rent out a domain in “clear”
محمد ثروت invited محمد ثروت
08:37
محمد ثروت
I Have a business proposal about creating an Arabic telegram community for $Amoveo
WHO should i pm?
Deleted invited Deleted Account
Deleted invited Deleted Account
16:18
Ric
No way! So cool.
16:18
Agreed having the dex would be great. This doesn't necessarily require
@zack_amoveo's involvement right? Anyone can grab the light node js, buy a domain and build a nice UI right?
16:36
x
you can easily generate an downloadalble app with with nativefier in a few minutes.
it will be faster than the server version
22:17
mr x
plz binary bet api to make silly little app to show people
18:12
Zack
Im looking into the binary bet tool idea.
The current binary tool is for making binary uniswap markets, so i am changing that one's name.
It seems like while making the crosschain swap tool, we created lamost everything we need to make a really good binary bet tool.
So the parts would be:
1) An interface to offer new bets. The offers are off-chain, and so unmatched offers are free.
2) an interface to view and accept bets from others.
When you are participating in a bet, it automatically makes an offer to sell for 99% of the max value. So you dont need to worry about making an oracle or enforcing anything.
23:09
Ben
please complete the dex first.
01:50
Webgru
Hi, who is admin?
02:02
Zack
http://159.89.87.58:8080/wallet.htmlI added the "bet" tab.
I haven't tested it much, and it doesn't have a button for canceling bet offers.
You can publish binary bets. you can see bets others published. you can accept them.
Not much new is here, I just copied tools from other places into an organized tab.
Anas Nour invited Anas Nour
09:23
Zack
I cleaned up the text in the new bet tab. Now it is easier to understand what you are betting on.
10:29
Toby Ganger
we desperately need a usable and functional product with a clean user interface….this four years of intellectual masterbation has gotten us nowhere….there’s a reason 99.9% of people don’t use command line on the internet
10:45
Zack
being able to make and accept bet offers is the most basic thing you can do with oracles.
It probably should have been the first thing we made.
At least it is working now.
10:46
it doesn't have these features yet.
a cancel button.
automatically try to sell you winnings for 99%, so you don't need to use the oracle.
17:51
mr x
how cancel button works? post tx that make nonce invalid?
17:53
Zack
When you make a swap offer, it is off-chain, and it has a number written to it so we can know if the swap offer is valid.
There is another number on-chain, as long as the on-chain number is lower than the number in the swap offer, then the swap offer is still valid.
There is another tx type that allows you to increase this number. So you can deactivate all the swap offers that reference that number.
17:54
Zack
For swap offers that allow partial-matching of the trade, each person who partially matches, they are only increasing the on-chain number a little. and once the on-chain number increases enough, then the swap offer deactivates.
17:54
the on-chain number, we decided to call it a "receipt"
17:55
there is a merkle tree dedicated to receipts now
17:55
We already have code for the cancel button in the DEX page. so we should be able to copy/paste it to this other tab.
17:56
mr x
ok. how about constantly refreshing offers 1 block forward and stopping refresh is canceling?
17:57
Zack
that would be a better way to do it if you could stay online.
You can re-publish an offer that expires in 1 block, but you need to keep clicking that button.
I could add another button that keeps re-publishing each time there is a new block.
17:57
mr x
just thinking about making a bot
17:57
Zack
this seems like a tool that we should abstract, so it can be usable in the DEX and in bets and anywhere else with swap offers
18:00
maybe it makes sense to do it in something other than javascript.
if they left the page, the thread would pause and stop publishing offers.
18:01
mr x
you can also leave the page open? xD
18:01
Zack
how about if the javascript light wallet generates the first swap offer, and you give that to an erlang program.
the erlang program keeps asking the network for the block height and checking if your offer was already accepted. and it re-publishes the offer at every block, after increasing the expiration and re-signing it.
18:02
if you click on another window in your OS, it might still pause the javascript?
18:02
It might depend on your OS
18:03
simple javascript bots to make it simple and accessible
18:04
automatically copy bets from other places type of stuff
18:06
Zack
My browser tabs can receive notifications when I am in a different tab.
I bet the javascript doesn't pause when you leave, ill look into it.
19:30
Zack
I added both of those.
Now it has cancel buttons, and it tries to automatically sell your winnings for 99% of their value.
20:44
mr x
doesn't seem to work
20:45
rpc.js:53 POST
http://0.0.0.0:8090/ net::ERR_ADDRESS_INVALID
(anonymous) @ rpc.js:53
atalk @ rpc.js:34
main2 @ rpc.js:31
doit @ bet_tab_builder.js:87
bet_tab_builder.js:100 Uncaught (in promise)
{status: 0, statusText: ""}
doit @ bet_tab_builder.js:100
async function (async)
doit
21:09
Zack
oh right. I forgot to switch back from test mode to production mode before pushing to the server.
It should be fixed now.
21:42
mr x
same error still
21:43
Zack
oh. I didn't type my password to actually push it to github
22:01
mr x
ok seems to work
22:04
"load an offer before you can accept it." when i try to accept
22:04
Zack
you need to open up the bet first.
22:04
first open the market to see the bets in that market. then open a bet.
22:13
Zack
did you see it say how many blocks until the offer expires?
22:13
I tried again now, and it seems to work on the test network
22:16
mr x
expires in 4 blocks.
22:28
Zack
I was able to accept both your swap offers
22:33
Zack
and my error for the second isn't the same as yours. it says I am sending a message to the wrong server.
22:39
mr x
now it accepted
22:57
Anas Nour
I was wondering if you have any note worthy competition?
22:58
For example could Ada use the harberger land registration to store its value?
23:03
Zack
bitcoin hivemind or gnosis seem like more serious competitors.
Amoveo is a blockchain for financial derivatives.
23:30
Anas Nour
Is the goal for amoveo to be able to store the value of properties and land?
23:30
Zack
That is a strategy we are looking at in order to have more liquidity for derivatives contracts
23:57
Eric Arsenault
I don't think people are ready for a Harberger model
23:58
I don't think I would use it personally for my property
23:58
imagine having to leave at any moment because someone buys your property?
23:58
Doesn't really make sense to me...
23:58
For gaming items: sure, risk is pretty small
00:03
Josh
You'd get a reasonable amount of time to move out and you can set the price to make it worth your while.
01:10
kartar
We have a team that work on making watchlist star and upvotes in any site like (coingeko,coinmarketcap,watcher,etc)
How can I dm to have a deal?
02:59
Zack
I doubt you need to leave in one moment.
Harberger is just a registry. If your location has a tradition of giving people 4 weeks to leave, then the harberger system will value the land slightly less, to make up for the 4 weeks that arent enforceable.
03:01
Dont you live somewhere with decent property rights though? If it is fairly easy for you to sell your property, then you are not going to be one of the first users of the harberger system.
Maybe people from your region would be more interested in holding veo, to profit off of other people paying taxes.
03:02
The users of the harberger registry dont need to be the same people who use the smart contracts.
05:05
Eric Arsenault
Yeah
15:56
Adam
Hello everyone
We have a team that work on making watchlist star and upvotes in any site like (coingeko,coinmarketcap,etc)
How can I dm to have a deal?
17:08
Zack
I was originally thinking that financial derivatives, contracts to trade risk, that they produce most of the value to sustain the market cap of the largest currency.
But now I think I had it backwards.
People want to hedge their risks in the least volatile and most liquid currency. That way they don't bring on new risks while trying to hedge old risks.
It seems like the market cap will tend towards ((the value burn rate)-(coins minting rate)) / (the market interest rate)
The burn can be in PoW or in tokens or anything with value that can be provably burned.
If we want to build the biggest blockchain for financial derivatives, we need to find a way for a lot of value to get burned. And we need to do it without imposing a cost on the holders of veo or the users of the financial derivatives system, and we cannot print new tokens to make it happen.
The only way some third party is going to be burning value for us is if they are somehow receiving something more valuable in exchange.
There are not a lot of mechanisms that fit all these constraints.
17:15
If there is limited access to a resource, then imposing a cost on users can be a net win for everyone.
Like in a crowded city, it is better to be required to pay for parking your car, because if parking is free, then you wont be able to find a place to park.
17:18
If you are shopping, and you want to pay for your goods and exit the store, it is better that there is a line. So you only wait about as long as everyone else needs to wait. People wont be struggling to cut in front of each other.
Needing to spend a little time waiting, it is a small cost that everyone is willing to pay, because playing this game is a net win for everyone vs the alternative.
17:26
I was in a fairly crowded city that had a lot of free parking.
The equilibrium they settled on is like this:
There are people who have cleaning tools, that watch each street. When you park, they offer to keep an eye on your car and clean it. If you don't pay, your car is at high risk of vandalism.
People accept the cleaner-thugs, because without them it would be impossible to find parking. They don't cost much, since they are all in competition with each other, you can park on the next street over. The existence of these thugs is a net win for all the players of the game.
17:32
Amoveo's oracle isn't able to know who is using a parking spot, or who is waiting their turn in line.
If we are going to meter access to a resource, it needs to be a resource that the blockchain can see the state of.
Land registries are the only thing I can think of that meet all these conditions.
The registry doesn't enforce anything. You are just paying to have your public key recorded in a database along with the coordinates of a plot of land.
17:44
Maybe domain names or brand names could work too.
💋💋💋 ГRUППА_ЗNAКOMCTВ_18+ 💋💋💋 invited 💋💋💋 ГRUППА_ЗNAКOMCTВ_18+ 💋💋💋
07:01
Toby Ganger
or I have a crazy idea….what if….we had a usable product? Like a product that a normal person can use…..or we can spend another 4 years theorizing about how our market is actually in tokenizing particles of air to revolutionize how people breathe…..
07:11
Zack
Did you try out the new bet tab?
07:12
Toby Ganger
the key words being “a normal person can use”
07:12
Živojin Mirić
Hi Toby ol fren
07:13
I would send a funny sticker of approval as a reply to your post but futarchy does not allow me
Jay Chan | IM community ☆ invited Jay Chan | IM community ☆
18:12
Jay Chan | IM community ☆
I from IM Community, i'd like to invite your project to hold AMA in our community, who can i discuss with?
18:23
Ben
the harberger stuff have so many reason why it will not work in real world. i don't see why you are wasting your time there.
18:24
Zack
can you share one of those reasons? the strongest one.
18:25
Ben
in germany authority is managing the lot's why would that ever change?
18:26
Zack
in germany it is fairly easy to buy and sell land and have an official title, as far as I know.
So Germany isn't a good place to implement Harberger.
18:27
Ben
all 1st world countrys are more or less like germany and even 2nd are the same.
18:28
and in 3rd the machine gun is the lot register.
18:29
Zack
There is a lot of space between Germany and machine guns.
There are places in the world where there exists a government land registry, but it doesn't work well. So there are secondary land registries that work in parallel.
18:30
Ben
so you are targeting a nieche market.
18:30
Zack
Many people already live with more than one land registry in parallel. and they are looking for any way to strengthen their claim to the plot of land.
18:30
Ben
i think investing your time into the DEX/Oracle could lead to a product that attracts ,Million of people
18:30
Zack
it is niche currently, if you look at in terms of the value of the land we are talking about.
But that land is lower in value exactly because of how weak the ownership claims are. No one wants to invest in land if their claim is weak.
18:31
Ben
no one wants to live in lands where is no economy and poverty
18:31
Zack
If these markets could let people have strong claims to their land, then we could see the value of this land ballooning to a majority.
18:31
You can grow food faster in places where the sun shines stronger.
18:31
x
i want but i cant find such land
18:33
x
land for farming, not in the city
18:36
and i no longer want to buy any real estate, because goverments are able to confisgate easily with just a decison. this problem cant be solved
18:36
Zack
If you compare the intense agriculture in Alberta or Saskatchewan, vs the kind of agriculture in Guatemala for example.
This part of Canada has a growing season of like half a year, and the sun is low on the horizon for all that time.
Crops are regularly destroyed by hail, even in summer.
Whereas in Guatemala, you can grow food all year long. The potential of the land is enormous in comparison, but no one wants to invest to be able to efficiently utilize that land. So it is full of subsistence farmers who gain like 1% of the possible utility of their land.
18:44
50% of Guatemalan jobs are in agriculture, but they can only satisfy 60% of their own food demand. It is ridiculously inefficient.
The tropics are like this all around the world.
They have the potential to produce food like 10x faster per hectare, but don't, because no one wants to invest in the land.
19:33
Zack
There are 3 kinds of dex so far.
1) veo to crypto.
2) crypto to veo.
3) crypto to crypto.
only (1) and (2) work so far.
(3) is a mixture of (1) and (2) together.
(1) and (2) already have a lot of duplicate code. It is pretty difficult to read.
Trying to write (3) directly by copy/pasting the parts I need is very difficult.
So what I am doing now, I am finding patterns in 1 and 2 so that I can rewrite them in a more readable and maintainable format.
Once 1 and 2 are readable, then it should be straightforward to combine them into 3.
22:52
Ilmu Somebody
settin' things straight! 😎
22:53
there are philosophical discussions to be had about land and how to allocate it but the fact is that amoveo has nowhere near the level of legitimacy it would need to enter that game yet
22:53
establishing amoveo as a solution to many current problems in the cryptocurrency ecosystem is a quick way to gain more legitimacy though
22:54
and I understand wanting to recharge and refocus a bit by thinking about next steps before refactoring everything and finalizing this step
22:55
but we need to figure out how to organize this UI business
22:55
maybe a document that can be given to a designer would be a good thing to coordinate on
22:56
explaining necessary fields and the order they need to be filled.. what they mean? that's a bit of a rabbithole but maybe clear references to well defined things and where to read more
22:58
Živojin Mirić
you can make UI and futarchy will decide it's worth
22:58
YOU HAVE BEEN FUTARCHIED, thx bye
22:59
Ilmu Somebody
if there was a UI and users then maybe that would be stating the obvious but right now it's a joke
00:04
Ben
the problem is the UI refresh does not make sens as long as the dex is not fully ready.
00:04
i would even fund a dev to create the UI
00:04
but currently it would be waste of time.
00:32
Ilmu Somebody
I very much disagree, if veo->alt and alt->veo is working then it is working enough. Problem with paying a dev or a designer is that the person wouldn't know what to do
00:46
Zack
how much legitimacy did bitcoin have in order to start playing the global currency game?
00:48
If something seems hard, that means less competitors will be willing to try to do it. If harberger land registries are going to be important, it is better that we get our version running sooner instead of later. First mover advantage means a lot.
01:26
Ilmu Somebody
🤷♂️ I don't disagree but so far nothing suggests that your method is working very well and although everyone just keeps whining and you are the only one doing anything I think you could do a better job using this community
01:26
despite that failing you are doing a great job and I've definitely learned a lot from studying your work
00:25
Jeans
I am not sure that it is easier to grow in a tropical climate compared to a temperate one.
But if you have any documentation on this, I'm interested.
00:31
Zack
look up calories per hectare of staple crops in the tropics.
kassava. sugar cane. palm oil. They are more productive than temperate crops like wheat or sugar beets or canola.
kassava in particular can make up a big component of animal feed.
When meat is created in temperate climates, they are feeding animals soybeans that they purchased from tropical climates.
But kassava is more productive than soybeans.
If meat is produced in tropical regions, the cost is lower, because they can feed the animals things like citrus and kassava, which are not exportable the same way that soybeans are.
Animals in tropical climates don't freeze to death the way they do in cold places.
01:01
mr x
trying to make bet
01:02
bet_tab_builder.js:100 Uncaught (in promise) {status: 0, statusText: ''}
01:02
Zack
ok, ill check it out now
01:07
Zack
I can't reproduce the error. but I have been fixing a lot of stuff like this lately. So I pushed my working code to github and updated the server. Hopefully it is working now.
01:08
ive been updating from callback format to async/await format.
It seems like this format is a lot easier to read and edit.
but if I forget to put an "await" in when updating the code, then this error happens.
01:13
Jeans
It's an interesting point of view, I'll read up on that. thank's !
01:23
mr x
still not working for me hmm
01:23
Zack
what step exactly is breaking? when you click "make bet"?
01:23
rpc.js:53 POST
http://0.0.0.0:8090/ net::ERR_ADDRESS_INVALID
(anonymous) @ rpc.js:53
atalk @ rpc.js:34
main2 @ rpc.js:31
doit @ bet_tab_builder.js:87
bet_tab_builder.js:100 Uncaught (in promise) {status: 0, statusText: ''}
01:24
Zack
oh. I left the code in test mode.
01:24
0.0.0.0 is for a node on the same computer
01:27
Zack
it should be working now
15:52
Godin*. Though not sure I agree with his PoS (standard) argument
19:17
Zack
Looks like the podcast is down.
Deleted invited Deleted Account
01:57
Zack
Ive got some much better tools for the dex now. veo->crypto and crypt->veo are both less than half as long as before, and far more readible.
It is a clearer what is going on in the crypto->crypto part now. There is a design decision we need to make.
The way the crypto->crypto tool works, in order to exchange 1 unit of BTC for 1 unit of ETH, there needs to be like 1.2 units of VEO locked up in the smart contract for the duration of the exchange, to enforce the different parts.
So the question is, who should provide that VEO? It is a pretty big trade off.
There are 2 participants in the DEX exchange. One of them (P) who posts a trade offer, and the other who accepts the offer (A).
If P needs to provide this liquidity, the process is shorter.
P posts the offer.
A accepts the offer and sends crypto1.
P sends crypto2.
A releases the VEO. (optional)
If A needs to provide the liquidity, it is longer:
P posts the offer.
A accepts the offer.
P sends crypto2.
A sends crypto1.
P releases the VEO. (optional)
I kind of like the version where P provides the liquidity, because it makes things very simple for A.
P is able to use the same VEO to post many offers, and if their VEO gets locked in some contracts, then the other swap offers all automatically become invalid.
So I imagine a liquidity provider using a small amount of VEO to offer user friendly exchanges between a lot of different currency pairs.
I think this is the version we had agreed on before, but I thought we could reconfirm that this is what we want.
02:00
The person who releases the VEO in the last step is necessarily not the person who needed to have more VEO in the beginning of the process.
02:05
What could be nice about A providing the liquidity is that P could make a bunch of offers, and then go offline. He doesn't have to worry that his veo will run out and cancel other offers, because it consumes very little veo.
It makes it so P doesn't need a hot wallet so badly, so it could be cheaper to participate as P and provide liquidity to the network without needing so much VEO exposure.
It is possible that people will want both strategies eventually.
04:09
Toby Ganger
cool! so investors won’t be down 95% in a bull market any more?
05:55
Z
This is great progress!
21:57
Ben
what about liquidity pools that provide the 1.2 ETH and in return get a fraction of the deal OR from the next block reward?
01:51
Karlis L
Sounds great, but it should not be the main approach, cause the functionality would depend on the existence of such pool + i assume a bunch of large outstanding orders would eat up all the liquidity of the pool and block trading for other users.
However, this could be a very nice extra service for people who do not have/dont want to hold any VEO.
01:55
Typically market makers (P in this case) are the ones with more available funds and they get paid via bid-ask spread for locking it in orders.
02:04
Ben
Liquidity pools are common, so no big deal. Whole defi rely on it.
06:15
Zack
ive been looking at the crypto<>crypto dex.
It looks like it is working for making offers. for canceling offers. and for accepting offers.
It still doesn't have the button for releasing the funds, and it doesn't have an indicator for where to send the btc/eth/whatever.
But it looks like we got past the part that we were stuck on. abstracting tools from the veo->crypto and crypto->veo tabs worked.
06:16
you can't store eth inside of amoveo.
We do have uniswap tools where you can provide liquidity to trade VEO and eth stablecoins.
06:17
Živojin Mirić
What is it that you are REALLY looking for?
06:17
THE TRUTH IS IN THE DEEP WATERS
06:18
Živojin Mirić
You know
06:18
You can try to escape the question but deep in your soul you know what I mean
06:18
An that's how I sleep at ease
06:23
Zack
Croatian music?
We've got to keep this channel focused on Amoveo.
06:23
Živojin Mirić
Nationallity is not a question
06:23
The truth is above nationallity
06:24
I dont care about that
06:24
You are missing the point based on national prejudice
06:24
Is futarchy THE TRUTH
06:25
Zack
I think google isn't good at translating croatian to english. I don't understand
06:25
Živojin Mirić
I am speaking ENGLISH
06:25
So what do you have to translate?
06:26
You have the ability to tap in to THE TRUTH
06:26
And you are wasting iz with irrelevant shit
06:27
Market cap is PROOF
06:27
Zack
we need the dex so people can onboard into veo
06:27
Živojin Mirić
Dex is ok but onboarding people....
06:27
How many years I am with you
06:28
I don't have any at all any incrntivr to manipulate to make money
06:28
Zack
prediction markets are only as accurate as the wisdom of the people who have easy access to participate
06:28
with the dex more people can more easily participate
06:28
Živojin Mirić
I don't own any veo but,, I am still here
06:29
THE TRUTH IS IN THE DEEP WATERS
06:29
You think you know shit, you have great potential but you still behave as naive pleb
06:29
I can't stand this anymore
06:30
Živojin Mirić
Živojin Mirić 19.09.2021 06:29:01
THE TRUTH IS IN THE DEEP WATERS
06:32
Živojin Mirić
You Zack are the definition of a COMFORT ZONE
06:32
You think you do shit bit you are faking yourself that you do shit at all
06:33
You can theorize about irrelevant shit all day
06:34
Zack
This update is changing over 600 lines, and I did it in a day.
06:38
Živojin Mirić
You are MASTER EVADER
06:39
You can try but you can't escape
06:40
Zack
Please don't spam the channel. lets keep this focused on Amoveo.
06:41
Živojin Mirić
I am completely focused on amoveo
06:42
Is not the futarchy to decide?
06:42
Zack
did you make a futarchy market about something?
06:42
Živojin Mirić
I don't know how to
06:43
Zack
use the "create futarchy" tab in wallet.html
06:43
Živojin Mirić
It's user unfriendly
06:43
Have you heard about UX?
06:43
Zack
you can ask questions here if it is confusing
06:43
Živojin Mirić
Many have asked
06:44
THE TRUTH IS IN THE DEEP WATERS
06:45
Your product is NOT USER FRIENDLY
06:45
And that's the main culprit of it's demise
06:45
You are living through a bull run without FRUITS
06:46
And your labour deserves fruit
06:46
I honestly think that
06:48
How many, years or bull runs are you willing to participate in
06:49
In which you have the potential to be THE TOP OF THE CROP
10:42
Zack
I ran into a problem in the tool for releasing the veo.
For this example Paul posted an offer to sell BTC for Eth, and Alice accepted the offer.
First off, the contracts.
We have a sell veo contract that divides veo into 2 types.
type 2 is valuable if the btc is delivered.
we use type 2 as collateral for the buy veo contract, which divides this collateral into 2 types.
type 2 is valuable if the eth is delivered.
So when Alice accepts the offer, she ends up holding buy contract type 1.
Paul is holding sell contract type 1 and buy contract type 2.
In order to settle the contract, we are going to need one of them to be holding all 3 share types to combine back to veo.
So one of then needs to sell their shares to the other.
But each of them is holding a share type that has lost. If they make an offer to sell losing shares, then it is cheap for an attacker to buy those losing shares and force us to use an oracle to get our veo out.
I think if Paul sends his losing shares to Alice for free at the right moment, then Alice will have an incentive to release the veo later.
P posts the offer.
A accepts it and sends BTC.
P sends the Eth and gives his losing shares to Alice.
Alice releases the veo by buying Paul's winning shares, and combining them with both types of losing shares.
Tersia Marais invited Tersia Marais
13:51
Toby Ganger
@ShakaRAMa he just ignores it everytime…
17:30
Zack
sending the losing shares to Alice is difficult, because it is difficult to be sure you have the correct address for Alice.
If you send the losing shares to the wrong place, then we need to use the oracle to get our money out.
Maybe we should just trust the explorer to store the correct tx history for our account, since at worst all they can do is make us use the oracle. and if an explorer is hosting bad data, we can all switch to using other explorers.
17:32
we can check if Alice actually has the other kind of losing shares in the correct balance. So if an explorer did this attack, they would need to lock up a lot of veo in contracts to be able to confuse us. So the attack can be more expensive than the amount of value it could destroy.
17:35
maybe it doesn't work. because the attacker only needs to leave his money locked up until we finish our dex swap. but our money is trapped for the entire 2 week oracle process.
17:47
Another option is like this.
P = Paul is publishing the dex exchange. A = Alice is accepting.
Paul publishes the offer.
Alice accepts and sends the BTC.
Paul sends the Eth.
Alice sends subcurrency buy_veo contract type 1 to Paul.
Paul combines all the subcurrencies back to veo.
What changed here is that now Alice is the one sending a subcurrency to Paul instead of the other way.
Alice can know who is on the other side of the swap, because she is the one who accepted Paul's offer. She signed a tx accepting that offer. She can verify her own signature, and remember not having signed over a swap_tx with anyone else that is using this same contract.
The problem with this strategy is that Alice doesn't have any incentive to click the release button to send the buy_veo to Paul. She receives nothing, and it even costs her a fee to do it. But if she doesn't do it, Paul needs to wait for an entire oracle process to get his VEO out of the contract.
17:52
How about this.
When paul sends the ETH, he should also make an offer to sell his losing shares for something slightly less than the cost of having money trapped in the oracle.
So if an attacker buys the losing shares, then the attacker is covering nearly all of the cost of having money stuck in the oracle.
Paul could discount his winning shares to make up the difference, so it isn't actually more expensive for Alice to buy both of the subcurrencies from Paul and combine back to veo.
Randy Beverly invited Randy Beverly
18:06
Zack
Gathering up subcurrency types to re-combine them into veo has that adiabatic feeling of programming a reversible computer, right?
It is like we are doing the protocol in reverse after seeing the result, to avoid leaking entropy.
18:26
Zack
I think Zivojin was asking about my motivations. with the "fruit of my labour" stuff.
What is important to me is that I do everything I can to try and build the next global reserve currency.
This means that I am sometimes making decisions that lower the odds that Amoveo will be only moderately successful, like Bitcoin or Ethereum are today. Because I am more focused on winning the bigger challenge.
It also means I am sometimes making decisions that optimize more for growing the market cap, instead of for growing the value of individual tokens.
23:24
Zack
I got the crypto<>crypto dex basically working.
I need to adjust some constants to make it incentive compatible, currently I think it is in Alice's incentive to not release the veo at the end.
Also some buttons aren't disappearing when they should, some other small changes for usability.
00:18
Zack
It is important that i update this to hide the second release button until after the first is pressed. Or someone could lose a lot of veo.
07:16
Zack
I did this security update, and made it remove unnecessary buttons.
It seems like some constants are still not right, because after doing a swap, the person who made the offer is richer by 10% the value in veo.
🔥🔥🔥 ПОIСK_ПARЫ_НА_ВЕЧЕР 🔥🔥🔥 invited 🔥🔥🔥 ПОIСK_ПARЫ_НА_ВЕЧЕР 🔥🔥🔥
17:29
Zack
I confirmed that the other 2 ways of using the DEX don't leave one person with too much veo at the end.
It only happens in the crypto<>crypto tool.
I think it has to do with there being 2 different offers, and they both need to be slightly over-collateralized.
17:54
Zack
I fixed the problem in the crypto<>crypto dex. now they both end up having the correct amounts of money at the end.
17:59
seems like the crypto<>crypto DEX works.
18:03
I think I want to add some security checks that the amounts of veo being swapped in these contracts is correct. that you aren't over-paying at any step of the process.
00:04
Zack
I added security checks to make sure the collateral is good.
And it tells you the implied value of veo in terms of whatever you are trading, to verify that there is enough collateral.
It also tells you how over-collateralized the veo is, to prevent free options.
00:05
It seems like the crypto->crypto dex is ready. If anyone wants to test it, go ahead.
Maybe I should do a little market making as a proof of concept.
00:08
Is there a market where we know there is demand for a dex, and I can verify that the existing dex have wide spread?
Like, what is the most profitable pair where it will be easiest to transition people to a new DEX?
00:08
maybe monero<>zcash, right?
00:48
Zack
It sure is a relief getting the crypto->crypto done.
I think the way we did it is great too. Lots of small reusable parts that will make it easier to build more dapps.
01:57
Ric
Amazing! Existing markets: the other day I wanted to swap ETH-SOL (or AVAX) and couldn’t find an efficient way. Probably will be some need for cross-chain transfers between ETH-compatible chains
02:05
Zack
Acting as bridges between the eth clone networks makes sense
02:33
Mr Flintstone
are there any particularly expensive bridges for something like native eth-> wrapped L1 eth for example?
02:34
nice thing about this is you dont need to worry so much about free option
02:35
could do btc <-> wbtc but not sure if fees are high enough to be competitive without a lot of volume
02:37
Eric Arsenault
Monero?
02:38
maybe we should focus on getting a lot of liquidity for a single thing working well
02:47
Ben
Typically it is hard to get stuff from BSC chain over
02:48
Without going thru binance
02:48
So what about eth (BEP20) to eth (erc20) ?
03:00
Eric Arsenault
these things are super easy and common now
03:00
anyways for example
03:31
Mr Flintstone
anyswap is 0.1%
03:34
i might be mistaken for their bridge
03:54
Mr Flintstone
might work given the binance bridge limit per address (?) i think is ~13 eth equivalent per day
Erin Fernandez invited Erin Fernandez
05:50
Zack
That is nice you found that out
19:06
Zack
looks like binance is charging $15 to get any tokens out of their network.
I wonder how many people are taking money out of binance frequently enough to be willing to learn the amoveo dex.
19:08
Maybe people who want to withdraw small quantities of crypto are the ones who would benefit most
19:13
looks like using the amoveo DEX for 10 veo worth of currency costs around 0.14 veo
so if you are trying to remove less than $1000 from binance, Amoveo's dex is cheaper than the binance bridge.
19:22
If I remove the dev tax, the cost of using amoveo DEX for a 10 veo exchange drops to 0.024 veo.
0.24% fee.
So that means if you are withdrawing less than around $60k from binance, amoveo's dex is cheaper. Which is bigger than the daily limit of Binance's bridge, so Amoveo's dex is always cheaper.
19:23
Maybe we should set the dev tax to zero for a while. At least until there is some users.
19:32
Karlis L
Who is locking the 1.2x VEO eventually?
Expecting random people to hold $1200 worth of VEO to save ~$10 in binance commissions is rather wishful
19:34
Zack
The way the dex is set up now, the person who posts the offer needs to lock up 1.1x veo. the person who accepts the offer needs to lock up 0.1x veo.
19:36
Karlis L
Ben's idea of some pool that's posting offers for some specific pairs seems pretty good. Especially if we start with something that does not require price discovery
19:37
0.1x VEO sounds reasonable for smaller quantities, but might be a dealbreaker for larger trades (at least in the beginning)
19:38
Zack
it is possible to build the reverse.
So the person who posts only needs 0.1x, and the person who accepts needs 1.1x
19:38
but having the poster lock up more is more efficient, because they can use their same veo to post many different offers.
19:40
Maybe the way to start is by finding a fiat based app that makes it possible to prove that you already sent the fiat.
So we could have a fiat onramp.
19:48
Instead of finding a way to save people $10 on their crosschain exchange, it could be better to try and find someone who is being totally excluded from a crosschain exchange. Like, their money is trapped in one system, and they can't get it out.
If we are the only option available, then people will be willing to try it out.
19:59
x
currenctly traders from China use otc (all big chinese exchanges offer this) to buy/sell crypto with fiat.. that includes otc trading between USDT and CNY.
That works quite well, and fast.
I just see dex as a safer option for trading, with exchanges i often worry about safety if I store too much funds for a long time.
19:59
Zack
International remittances are a big pain point in the financial system, and so are intra-national remittances.
In a lot of parts of the world, the people don't have good access to banks. It can be expensive just to travel to where you can visit a bank.
A lot of people are depending on these phone based apps, where your cell phone company will let you send like $10 or $20 to your family who lives outside the city, and fees can be over 10%.
10% of $20 is only $2, which is cheaper than the cost of traveling to a bank for many people.
When they receive the money in their phone, they can visit some local shop the sells vegetables and eggs or whatever. and the shop keeper is a part of the cell phone network, so you can receive your $20 payment there.
It would be nice if we could find a way for these people to use VEO for this instead of a cell phone company. 10% fees are really high.
We don't need a full on crypto-ATM for this application, because it is very small amounts of money, and the shop keep is already providing this service.
20:00
safety is a good point. Maybe that is what we should focus on.
20:58
Mr Flintstone
the 0.1x is also arbitrary
21:20
Zack
Not arbitrary.
It is a free option protection.
If the veo/other currency price shifts by more than that much, the trade becomes undercollaterolized, and they can steal some money by using the oracle instead of settling normally.
21:21
Mr Flintstone
maybe arbitrary is the wrong word but the amount of money you need to secure 1 hour of volatility isnt 10% of the value of what is being traded most likely
21:23
and it also scales with the amount of time you decide the trade will take so can be even less
21:25
stealing money in this context seems like its basically having your order slip a bit
21:25
so the overall execution cost is the weighted average of all the outcomes which still can be rly low with some slippage events
21:26
Zack
Is it one hour though?
Like, what if i wait around until the oracle is half done, and then click the "release veo" button?
21:27
Maybe we need some time limit, after which the oracle is mandatory
21:28
Mr Flintstone
the trade is finished as soon as the coins on the other chain are or arent sent so everything can be settled among arbitrageurs aka ppl willing to wait a week and the ppl who need liquidity now if there is any funny business right?
21:28
Zack
Oh right. There is a date by which the money needs to arrive
21:29
Mr Flintstone
so you just make the time limit customized to the slowest/highest variance chain you are using
21:29
Zack
So maybe we can get it down to 1%, especially if veo was less volatile, or we used a stablecoin as collateral instead of veo.
21:29
Mr Flintstone
and then set it as low as you want
22:18
Ben
the stablecoin woule make a lot of sense actually
22:19
gives the user full transparency and creeates trust
23:59
Mr Flintstone
wonder if it is worth the effort to do retroactive incentives for providing unmatched orders i.e. ppl providing liquidity for stuff like eth <-> bsc eth will get a proportional share of x veo per day in the future
Bit Guru-10 invited Bit Guru-10
00:34
Bit Guru-10
Hello
I have a deep knowledge of blockchain technologies, specifically in DEFI, DEX, NFT projects so far. Build Smart Contracts using Solidity and deploy them to Ethereum and BSC using HardHat and Truffle, and interact web3.js/ethers.js, React.Js, Vue Js with Smart Contracts and more with industry basis. I’m full-time available and able to commit the work immediately.
Do you have any open positions that I can help you in?
Thanks!
00:37
the offers are off-chain though. How can we be sure that they existed?
You can create the invalid offers after the fact and send them to the oracle reporters.
I think if we did this, it wouldn't be possible for me to prove that I am not cheating the system to get free veo.
00:53
Mr Flintstone
if a market says some semi centralized scheme will increase the price of veo then idk
08:08
Eric Arsenault
Totally
08:08
We should really have incentives for this thing to bootstrap liquidity
08:11
Figuring out the incentive game here will be a big unlock
22:19
x
the project itself need more marketing, e.g. ask people to tweet about something and get small reward..., i saw a few projects did that, they even programmed that into the wallet.
also translation of content
maybe the oracle can reward people small amount of veo for any kind of referal acitivity
22:21
Z
Better to have the DEX working and UI all nice and done before doing marketing
22:22
K
I'm genuinely curious as to why you dont hire someone to make a front end for amoveo tools?
22:27
x
https://myveowallet.com/ some people did that , the traditional way. probably they are not working on it currently
this has not been my concern.
I always believe the product is ready, but marketing is lacking( for example introducing veo to more countries, currently it is known mostly in the english market). dex is just another feature.
22:31
Zack
Maybe at this point we should keep trying to use amoveo to do fun things.
The bet offer interface seems to have good potential.
Maybe we can make some attention grabbing futarchy.
Futarchy markets about other crypto projects. I think we did that before and it got good attention.
22:32
If we just keep using it until it becomes a smooth enough process to be enjoyable
22:33
x
satoshidice was a very simple betting game in the early days of bitcoin, they don't have a ton of feature, but was able to become the most popular app.
22:44
The DEX can become more and more popular even without ux, centralized exchange will have kyc, AML, more strict withdrawal policy etc.
22:56
K
Aren't there more features out now that this wallet doesnt cover?
23:00
Zack
Yeah, my veo wallet only works for spending and receiving. It doesnt do any smart contracts.
02:43
Eric Arsenault
Has anyone swapped from one blockchain to another yet?
02:54
Zack
I think not.
Bitcoin tx fees are low now
Ladas Run invited Ladas Run
💥💥💥 PЕАЛЬHЫE_СЕКС_3НАК0MSТVA 💥💥💥 invited 💥💥💥 PЕАЛЬHЫE_СЕКС_3НАК0MSТVA 💥💥💥
Shivan Group invited Shivan Group
Faux Kreso invited Faux Kreso
BitKan Nicole invited BitKan Nicole
AMAN SHARMA invited AMAN SHARMA
16:05
AMAN SHARMA
Hey Admin i have found vulnerablity on your main website how should I report it??
16:05
I'm a cybersecurity student
16:06
Zack
I sent a direct message.
15:01
mr x
when you make 99% offers they also expire at same time as original offer even though shouldn't?
15:02
Zack
Right, that needs to be fixed
17:09
bullish for all DEXs?
01:11
Eric Arsenault
Yuuuup
Luca San joined group by link from Group
Deleted invited Deleted Account
Epsita | Hyper Growth invited Epsita | Hyper Growth
23:27
Epsita | Hyper Growth
Hey there Team, I hope you are doing well!
I am Epsita from HYPERGROWTH Marketing Team. We have developed some unique Marketing Techniques and we are working with a major Launchpad and many projects and they all are happy with it.
I would like to propose the same to your Project and help you and reach heights together.
Who should I contact for Marketing Proposal?
01:00
Z
This is amazing i love it
Deleted invited Deleted Account
17:26
Ric
+1
PS why is futarchy not more widespread? Just not well implemented anywhere? Or is there some inherent awkward UX?
17:56
Zack
The shortcoming that surprised me.
If one outcome is much more likely than the other, you cant get a good reading from futarchy, and it is cheap to manipulate the price to make the less likely option look likely.
17:57
Zack
But i think we might find tricks to overcome this, with experience
18:20
x
I'd be suprised if there are more people using the futarchy
0 effort to teach people how to use it, or to make it so easy to use that it require no teaching...
Anyway, amoveo can do a lot of things,
You need to start somewhere,
I do see prediction markets, and DEXes being promoted on other blockchains.
18:58
x
it's not about awkwardness, it does not have to be beautiful. but easy to learn and use is important.
Brian_Epid Team invited Brian_Epid Team
Brian_Epid Team invited Brian_Epid Team
Semilogo invited Semilogo
Crypto Boss invited Crypto Boss
01:29
Zack
facebook is down for political reasons, right?
So we could bet on how long it will stay down.
01:29
Živojin Mirić
No it's not
03:30
O K
Is it? What happened?
03:31
Ben
Someone removed the BGP routes from the Facebook peering router via a web interface
03:32
After they recognized the failure it was already to late and the cut themself out from the system
03:32
Now they need people that go to the router and have physical access
03:33
Murphy did a damn good job today :)
20:40
it is still a really rough idea. I need to compare more alternatives and more formally calculate the incentives of participants.
20:46
Živojin Mirić
is your implementation of futarchy still based on the pure irrational will and personal interests of the participants as before?
21:00
Zack
yeah, maybe even more so than before.
But mostly I am just trying to find a way to make it secure against manipulation, for those cases where our current design is vulnerable.
21:01
Prediction market futarchy needs a metric to optimize for. like hashrate.
risk hedging futarchy doesn't have any metric. Each person is optimizing based on their own personal values.
So it is even more closely aligned with the interests of the participants.
21:06
I think maybe the risk hedging futarchy doesn't work, because of the embedded minority wins game.
People will only participate as long as they think that they are on the minority side.
21:39
Zack
yes, this is about the same idea
21:41
I was wrong in our discussion in DcInsiders. People can manipulate the low-probability side of the market in either direction, and hedging against the high-probability outcome is infeasible because of how much money you need to lock up, and how much you will lose because of the interest rate.
21:42
Maybe we need to modify how futarchy works, or have several different futarchy mechanisms that are optimized for the different situations.
Like, it seems that our current futarchy mechanism works very well for the Amoveo community, since we are dedicated to following the advice of the futarchy.
22:08
O K
Maybe we should bet on which form of futarchy is the most feasible
22:12
Zack
Haha
That might trigger the singularity.
Like a self improving ai.
22:44
mr x
blockchains now go hashrate -> money
22:44
money must point back to hashrate
22:44
complete the circle
01:07
Zack
Thinking more about risk-hedging type futarchy.
It is like the people who are in the majority get what they want, but they need to pay compensation to the people in the minority for damages.
Maybe the way to do it is that the people who vote with the majority, they lose only ~3% of the money that they had used to vote with.
And people in the minority, they get their money back, plus that extra 3%. but since they are a smaller group, it is like they get their money back and 5% interest.
so this is like a linear combination of the pure risk-hedge futarchy idea along with letting people vote based on how many coins they own.
01:21
Another option is a combination of the risk-hedge futarchy idea with the prediction market futarchy idea.
We use the 4-outcome combinatorial market from the prediction market plan, but we interpret the prices differently.
Instead of going with the strategy that optimizes for hashrate or whatever metric we care about, we optimize for some combination of hashrate along with preferring whichever decision is the minority of bets.
So if the prices are like:
1 0
0 1
Where most bets are along a diagonal, then it is working like a prediction market. it is clear one outcome is better for society. We pay out bets normally.
But if the prices are like:
100 0
99 1
Where most of the bets are on one side, then it is working like a risk hedging market. More people want to hedge against one result than the other, so we can tell that more people would benefit from one result than the other according to their individual values.
In this case, the decision that the market says wont happen, we go with that decision. But, we pay out bets a little differently.
People in the minority who win the market, they only win like 10% more than they had bet. And all the people in the majority, they only lose a small fraction of their money, they get refunded almost everything.
Deleted invited Deleted Account
01:33
Deleted Account
Hi buddy
01:34
Please who can I contact for a professional proposal of C.M.C. listings within 1-6 hours! Thanking you in anticipation for your quick response.
01:39
Deleted Account
Thanks
18:02
Zack
There are a couple ways I can imagine futarchy winning.
The fast way is that we start using futarchy to show how broken the current methods of decision making are. People realize that their politicians and bosses are making terrible decisions, and demand to start using futarchy.
But it seems like futarchy might only work for communities that are already dedicated to following the decision of the futarchy.
It is like Tinkerbell, she can only let you fly if you believe that she will let you fly.
If futarchy only works for true believers, then we can't use it to show how broken the current methods of decision making are.
In that case, the rise of futarchy might follow a different much slower path.
A few small communities of believers start using futarchy to make decisions, and these communities are more successful than the alternatives.
They attract new members more quickly, they earn profit more quickly, etc.
So, after enough time, the dominant communities are all using futarchy.
If we follow the slow path, it seems like it could take generations for futarchy to win.
18:04
Živojin Mirić
when it gets big enough it will attract saboteurs and manipulative malicious actors, that's the human nature and it's not compatible with any idealistic ideology
18:08
Ilmu Somebody
Burn the infidels!
20:53
Zack
If futarchy only works for communities that believe in it, then maybe it will start to seem more like a religion
21:12
Ilmu Somebody
it is. We are trying to stave of the forces of Isfet, Ma'at's protection wears thin 😁👾
07:21
Zack
I did
mr 's idea so that 99% offers expire later.
08:03
Zack
I updated the bet tab to use the abstracted function for 99% offers.
I think it solved the same bug in regard to expiring later, and it fixed a bug where it was trading at the wrong price.
Tomasz Wojewoda invited Tomasz Wojewoda
12:48
mr x
nice. 12000 like 90 days
12:50
offerpool gonna start bloating? :P
13:04
Eric Arsenault
Can an external community use Futarchy on the Amoveo blockchain somehow?
13:05
We have a DAO, and constantly need to make decisions on funding people: how can futarchy help?
13:05
We want to optimize for transactions in our protocol
13:24
Toby Ganger
at least a religion has a community
16:50
Zack
They can. You can decide who to fund and optimize for number of transactions yes. That sounds like a publicly measurable metric that the oracles could report on.
17:17
Zack
I heard how something like 20% of people in USA are functionally illiterate, and it is similar for developed nations.
It seems like the number is even higher in countries that don't have as good of an education system.
If we made an app that targeted illiterate people, it seems like that app would also work for people who don't speak english.
It would give us access to an entire demographic that other blockchains are all excluding.
Betting seems like it requires reading. you need to know what you are betting on.
But maybe the bet can be read out loud?
The DEX seems like it could work with lots of logos for different currencies, and have decent default settings.
19:18
Živojin Mirić
When I'm drunk I'm telling the truth!
19:19
Instinct
Haha u sold your veo?
19:19
U will be left on the sidelines when the day of reckoning comes
19:20
Živojin Mirić
I thought I have a backup wallet but it says 0VEO
19:20
Živojin Mirić
Wasn't much so not a big loss
23:16
Ilmu Somebody
you dont think this is a community? this channel is pretty active in a good way
03:05
Eric Arsenault
Our situation is that we have a DAO, with a pool of funds. And we are deciding as a group how to spend those funds...so it’s a bit different, not sure if futarchy can be applied to a DAO with a treasury
03:13
Zack
For futarchy to work, the people who are deciding as a group, they need to come up with some metric that they all want to optimize for.
Like, if the DAO has some currency you want to increase the value of, or if you want more money deposited in the DAO.
And the metric you optimize for, it needs to be something public that Amoveo oracle reporters can look up.
Also, which decision your DAO makes needs to be public info that the oracle reporters can look up.
05:48
Eric Arsenault
I guess people also need to put their money on the line
05:49
In our case, that’s not true
13:56
Zack
Futarchy can still be applicable.
Why they don't want to make reports?
Maybe because they don't expect to earn anything.
Maybe because they don't yet know the consequences of the decision, and it would cost them something to find out.
Maybe because it costs something to figure out how to use the futarchy tool and make a report.
The solution is to pay them more.
In prediction markets, the liquidity provider and the reporters are different participants.
The liquidity provider is basically paying reporters who make accurate reports.
For example, in the LMSR design for prediction markets, the liquidity provider has an explicit expectation of losing almost 100% of the liquidity provided.
The way this traditionally was expected to work, like in Robin Hanson's papers, is that a company that wants to make a decision, they provide the liquidity for the prediction market. They take an expected loss, as a way to pay the market for information that they need.
So for your case, maybe the DAO should use some of it's funds to provide the liquidity. You can see that Amoveo's futarchy tool lets you decide how much liquidity to provide in the constant product markets.
This is how the liquidity provider decides how much to pay for the information.
Deleted invited Deleted Account
Mugesh M invited Mugesh M
00:08
x
what does that mean? why isn't that ended? I clicked at some of the buttons to see what will happen not sure if i clicked on that one.
00:10
Zack
looks like the oracle was finalized
00:11
so, was the contract finalized too?
00:12
I guess not, because this page doesn't say anything about it being finalized
00:15
oh, looks like it was closed. ill update the contract explorer to handle this case.
00:16
so I guess you just need to withdraw your winnings, if you had participated.
00:16
and the reason the price is weird is because people withdrew their winnings in an odd order
00:20
maybe you need to take the money out of the constant product market first, and then you can convert it to veo
00:22
x
I didn't participate. I saw that list of contract today and thought they are expired but still there
00:23
Zack
maybe it is my money, haha
00:26
looks like it isn't mine. The pool tab isn't letting me sell any liquidity shares from this market.
00:32
mr x
not me i think
00:33
think i created the market and made couple veo lol
00:34
Zack
it was probably the most popular one so far.
00:34
mr x
peak liquidity 8 or something haha
08:15
Zack
I did a big update to the light node.
I switched a lot of callback syntax to await/async syntax, which is easier to read and maintain.
I updated the http post library to a newer more stable version, which I think is solving a memory leak as well.
This probably introduced some bugs.
It looks like the futarchy tab is not working.
The other tools seem ok.
I will get futarchy working again soon.
20:15
Zack
I got the futarchy tool working again. looks like it was a rounding error that made a balance go negative.
I also set up the merkle proof verification function to not use callback syntax. to help the code be easier to read and maintain.
02:07
Zack
in the light node, you can see there is a lot of javascript in this one file.
https://github.com/zack-bitcoin/light-node-amoveo/tree/master/src/jsSo, I am working on a plan to organize it into files.
I already moved the crypto related pages into a file crypto
and it is working with the version served from the full node as well as the stand alone version.
02:27
mr x
yeah good idea
02:32
should remove all ui stuff in crypto folder?
02:41
Zack
getting UI stuff out of the headers and keys library might be a good idea.
But I am going for more of a ravioli organization, with lots of little modules instead of the controller/model/view organization you see more often in javascript.
When it comes to verifying the security of code involving setting up your private key, sometimes it makes sense to have it all in one place.
seperating the interface for headers is probably a good idea.
08:13
Zack
I set up a meeting with a partial homomorphic encryption expert. He is going to help find out if there is some way new kinds of cryptography can help amoveo.
08:16
is there any blockchain cryptography system we would want him to be aware of?
08:18
proof of succinct work is a very reasonable upgrade to proof of work
麒麟银联䒫収商务 invited 麒麟银联䒫収商务
00:59
Zack
Im looking into partially homomorphic cryptography.
What if we find a way to make oracles more private.
So a smart contract's result depends on an oracle. But third parties who look at the smart contract, they can't tell which of the oracles was used to enforce this smart contract.
I am not sure if this is possible, but if it is, it seems like something we would need.
01:14
Mr Flintstone
couldnt you do it with zero knowledge proofs
01:15
Zack
yeah, maybe.
The question is more about whether we can do it at a low enough cost.
Like, we could put the entire SHA256 algorithm inside of a fully homomorphic crypto system, and verify the merkle proof in a zero knowledge manner, but that would probably be way too expensive.
01:18
one potential strategy is the ZK snarks like they are investigating in ethereum. so the computation happens off-chain, and there is a shorter verification that happens on-chain.
But maybe we can use a faster partially homomorphic system. converting the merkle proof into some kind of computation over a group built on an elliptic curve.
07:42
Zack
They suggested looking into "secure pattern matching", as a way to read from the database without revealing which entry in the database we are reading from.
08:00
Zack
They say that we can do efficient partially homomorphic encryption with XOR, and addition. and we can make it possible to know if a given encrypted value is a zero.
They want to know if there is anything useful they can solve with these tools.
08:21
Ilmu Somebody
You could count up to a threshold
08:22
Zack
it seems like with secure pattern matching systems, we would need to read from multiple locations in the oracle database, and people watching us, they wouldn't be able to know which of the oracles we read was the one that enforced the outcome.
08:29
Hopefully we don't need merkle proofs for a bunch of different oracles.
18:02
Zack
Maybe the way to do this is to have a mandatory cache of oracle data.
Every full node is required to keep in ram something like 1000 oracle results (~33 kilobytes of data)
And oracles can pay more to be kept in the cache longer.
So as long as the oracle data is in this cache, we could quickly run the secure pattern matching protocol on that data, and no one can tell which oracle your smart contract is reading from.
18:17
to be able to process blocks in any order, we can't depend on state from a cache like this.
18:19
We could write the entire 33 kilobytes directly in the block.
18:22
It seems like it should be possible to make something like a merkle proof, but all inside of homomorphic encryption.
So you can prove that something is in a database at a location that you are searching from, without revealing where it is that you searched.
19:13
Zack
Maybe we can assign oracles to the integers as they get created.
RSA accumulators store a value for each prime number below some limit. They start out empty.
For each integer we assign 2 adjacent prime numbers.
1 -> 2, 3
2 -> 5, 7
3 -> 11, 13
So each oracle, it has 2 locations in the accumulator. That way we can store the 4 possible states of an oracle: unsettled, true, false, bad_question
So if you want to reference an oracle, and you already know the integer that was assigned to that oracle when it was created, then you can check whether the 2 prime numbers related to that oracle are inside the accumulator or not.
19:14
This kind of privacy preserving smart contract could only be written after the oracle was created on-chain.
So we can't do that trick where we make the oracle at the last minute to customize the wording.
19:15
It looks like zcash is using a zero knowledge proof that an unspent transaction output is inside of a big database.
We could probably use something similar to make a proof that an oracle with certain state is in the database.
19:44
Zack
maybe we can even reuse some of the zcash codebase
19:54
oh, we can't generate the oracle question during the smart contract. that breaks privacy the same way.
So we would need the oracle to have existed since before we made the smart contract. We can't do last minute customizations.
19:54
but that oracle could potentially reference a hash of the actual text.
So the oracle text isn't revealed unless it becomes necessary for enforcement.
19:55
so maybe RSA accumulators are best.
22:07
Zack
the chronological perspective.
a pair of people decide that they want to make a secret bet.
They make an oracle that has a hash H for it's oracle text.
They make a smart contract that points to the oracle.
The loser refuses to end the contract fairly, so the winner is makes an oracle report, so the oracle will resolve correctly and they can win.
The loser makes a lying oracle report with more money than the winner has.
The winner makes a big honest oracle report, so they can double all their money. Then the winner publishes the oracle text that hashes to H, so that other people can make oracle reports as well.
The winner is only going to need to reveal the oracle text if:
1) the loser has refused to release the winnings to the winner.
2) someone is making large dishonest oracle reports.
With the private RSA accumulator, maybe even after the oracle text gets revealed, your smart contract is still not connected to that oracle text.
So what im wondering is, how meaningful is the benefit of this extra privacy provided by the RSA accumulator? It seems like only in very rare situations would it be needed or used.
And if someone is already losing money to make dishonest oracle reports, wouldn't that person also be willing to reveal his RSA secrets and let 3rd parties see what it was that you won your bet on?
22:18
Disconnecting the money from the text seems like a move in the right direction.
It keeps the money fungible too. By preventing it from having a history.
22:32
How about this.
If a bet is connected to it's hidden oracle, and the bet is for a very large amount of money, I could imagine that people are curious.
They might make some false oracle reports just to try and find out what the bet is about, or to bother the person trying to get their money out.
If the bet and it's oracle are disconnected, then people who see the large-value bet, they can't tell which oracle it is connected to.
22:38
Another benefit is that you can reference an oracle without needing the merkle proof. So the transaction is smaller.
19:15
Zack
It looks like RSA accumulators are no good.
They depend on a trusted administrator, who, if compromised, would be able to generate zero knowledge proofs that look valid, but are lying.
19:15
merkle tree based accumulators might work.
19:24
Ilmu Somebody
This is amazing
19:27
Zack
https://zkproof.org/2020/02/27/zkp-set-membership/Maybe im not understanding the trusted administrator part. This other page really makes it seem like RSA accumulators are the way to go.
I think ill start by implementing a non-private rsa accumulator, and then try and figure out how to make it private.
20:39
Zack
Looking at other use-cases for RSA accumulators.
https://ethresear.ch/t/open-problem-ideal-vector-commitment/7421It seems like vitalik is saying that RSA accumulators would work for vector commitments, as long as we never delete anything.
Our immutible contract system has this property.
Oracles only resolve once, so they have this property too.
It seems like if we used unspent transactions instead of accounts, that we could have 2 database that we never delete from. one with all transaction outputs, and another with all the spent transaction outputs.
20:45
oh, I see the problem.
proofs of membership are only valid for a single vector root.
So we need to validate against old roots sometimes.
but an old root wont know if a member has been removed.
Haroon anmol invited Haroon anmol
20:47
Zack
Currently merkle proofs make up around 90% of each block.
If we switch to using RSA accumulator proofs, each proof is fixed size, and very small. something like 64 bytes.
Sergey Balashov invited Sergey Balashov
22:07
Sergey Balashov
Hi, please config not connected, node
22:37
Zack
@Simon3456 i think he might be asking about how to connect to the mining pool
22:41
It doesnt allow for proofs of non-existence.
So if someone tries spending from an output, they wouldnt be able to prove it hadnt been spent before.
So maybe it doesnt work for balances.
22:43
Rsa accumulators seem impossible.
How can we fit do much information into one number?
Doesnt that break information theory somehow?
22:48
It seems like we can store 1 bit for every prime number between 2 and 2^256.
And the commitments and proofs are all only 512 bits.
So, could i encode 2^256/(256* log_e (256)) bits in only 512 bits?
Thats like, more than all the information on the internet. Maybe more than the info in the universe.
23:27
Sy
90% want to join with their BTC asic... 😔
23:42
Zack
I think what is going on here is that the information is not in the filter. It is in the query data.
00:21
Sergey Balashov
Help me pool settings please
00:22
Zack
What mining hardware are you using?
00:26
Sergey Balashov
Sha256 s9
00:26
Zack
That is for the bitcoin network. you cannot use that to mine for Amoveo.
00:28
Sergey Balashov
and what can be used for Amoveo
00:29
How to algoritm Amoveo mining?
02:41
Zack
It seems like for RSA accumulators, generating a proof that an element is in the set, it takes O(number of elements in the set) computations.
02:59
Z
Hello, i have some FPGA for sale if you want you can PM me
Type of machine: 3x Blackminer F2
Reason for sale: upgrading to L7 and power grid won’t support both
(Will only post this once)
04:57
Zack
another limitation is that checking if an element is in the set, it could be linked to an old root.
So you need a way to look up old roots on old blocks.
So this is adding memory for every block where new elements are added.
05:00
maybe the way to do rsa accumulator privacy is that a trusted third party could build up the rsa accumulator, and help us to generate zero knowledge proofs of the data inside of it.
So we could bet on anything in that trusted third party's accumulator, and they don't know what we are betting on. They don't know which bets are connected to which oracles.
09:39
Zack
RSA accumulators seem like they would be great for verkle trees.
If a node has 100 children, you convert take the 32 byte hashes of each child and convert them all to primes, then store those.
So you could have verkle proofs of length 64 bytes*log100 (# elements) instead of the current 32 bytes* log2 (# elements)
The limitation is that you couldnt make a proof of non-existence. But we dont need non-existence for enforcing oracle outcomes.
09:41
We use proofs of non-existence for accounts. Before creating a new account.
I wonder if we use them for the other merkle trees.
09:53
looks like almost all the trees have proofs of non-existence.
only governance, matched, and unmatched dont.
But we could make new data structures for parts of the information that we want more accessible. Like the outcomes of oracles.
10:33
Zack
How about if every block at first contained the merkle proofs, the merkle roots, and some rsa roots.
After the block already exists and is being gossiped, we can start building the expensive rsa proofs, and putting them in place instead of the merkle roots.
And the rsa proofs are gossiped. Rsa proofs are muvh shorter and easier to vetify. So it is like, all the blocks are shrinking all the time.
This way miners can start mining immediately on a large merkle proofed block. And once they find a solution, they can immediately send it to other mining pools.
Then lower resource nodes can sync at their own pace. Waiting for shorter versions of blocks when necessary.
There are certain nodes that specialize in making rsa proofs. The storage cost is O ((# blocks you can revert)*(# elements in the accumulator))
The cost of building the next version of the database after a block is O (# elements in the tree)
The cost of each rsa proof generated is O (# elements in the tree)
10:54
Zack
If every person had an address, that is around 33 steps in a binary merkle tree. 512 bytes per step. Around 16.5 kilobytes per account referenced.
If we used a rsa accumulated verkle tree with 2048 children per node, then the proof is only 3 steps, 512 bytes per step. Around 1.5 kilobytes per account referenced.
Blocks would be almost 10x shorter after compression.
10:58
We only check if an account is empty when we are creating a new account.
So we only need a full merkle proof when creating a new account. The rest of the time, we can compress to an rsa accumulator verkle proof.
12:08
Shivan Group
No one is sharing fpga bitstreams…. Sooooo no one can mine the coin unfortunately. Asked Sy on discord. Great project but if no one can mine it’s DOA my friend. *sad horn. Unless the bitstreams are shared.
12:09
Ran gpu and it can’t keep up with the bitreams…. So more DOA
12:09
One person is mining everything… lame
12:11
Someone had to say it
12:49
Shivan Group
The mining channel was “less than friendly” and had more of a “go way don’t bother us” discussion. Makes me think the bitstreams aren’t being shared deliberately by a few. And I’m not new at this.
12:49
Ben
Just buy some F1+ blackminer and you can mine it
12:50
Blackminer never disclosed any bitstreams, its their business strategy and has not to much to do with amoveo
12:51
Shivan Group
You need the right bitstream.
12:52
Ben
It is included in the blackminer dude
12:53
Shivan Group
That’s the only one you can use then, and not load others via xlinix
12:53
That’s the designer of the coins intent?
12:54
Look at the veopool dude
12:55
Ben
Sounds a bit like you are late to the party and now you are blaming everyone who is already drunk
12:58
Shivan Group
Hah, one user has the 50% the pool. It’s not a party, it someone drinking alone watching a coin fail
12:59
Ben
This coin is not failing due to mining
12:59
Shivan Group
Network hash is ~2400 gh/s and one user has ~1200 gh/s and the admin of discord tells everyone to use gpu
12:59
Ben
It has other more serious problems
13:00
You can buy a blackminer and compete
13:00
one is doing 80 gig
13:01
Shivan Group
Fundamentally this is why fpga is going the way of the dodo.
13:02
Economically restasking other miners with a bitstream will enable the market independent of other “issues” I’m sure are there
13:04
Someone with 15 black miners is hitting every block. So the entire network is a couple people with maybe 25 black miners.
13:06
There is clearly demand, sy’s comment above is that 90% are turned away.
13:07
This coin should partner with black miner if they want them to control supply and devalue all other existing veo coins, or release the bitstream
13:26
Ben
again, this coin has fundamental problems, but mining is non of it.
13:33
Shivan Group
Dunno, you could have the best coin on the planet. Perfect code, perfect everything. If no one can access it, then it will fail anyway.
14:34
every aspect of Amoveo that’s human-facing is a disaster
15:46
Ben
but the game theory! Oh Boy
16:45
Zack
It is single sha 256.
We have open source in C, and for gpu.
How do you make a bitstream?
16:45
If we had a bitcoin bitstream, we just remove one of the 2 hashes
17:50
Zack
Because of oracle game theory in amoveo, almost none of the oracles ever get finalized.
Participants in a smart contract that could potentially need the oracle, they have an incentive to cooperate to end the contract at the correct final state, without needing to wait for the 2 week oracle process to complete.
So if we were storing oracle results in a rsa accumulator, we would need to reserve 1 prime number for every oracle created. But the computation only gets expensive for every additional oracle that was resolved.
So one strategy we could use, we could have 2 rsa roots. Each one is keeping track of some of the 10 000 most recently finalized oracles.
When one has 10k elements, it deletes it all and starts over with 0 elements. And at that point, the other one has 5k elements.
So there is always access to proofs for the 5k most recently finalized oracles.
The cost of generating a proof is something like 10k multiplications, inside of a finite field.
It is cheaper if you make proofs for multiple things at the same time.
Verifying proofs is constant time. Proof length is like 512 bytes.
Vs our merkle strategy. Where the proof length for 10k elements is 13*512. So 13x worse.
And this rsa strategy might have privacy for which value we are accessing.
17:50
Tiger Dev
Hi All
I am a Senior Blockchain and Full Stack Developer with core expertise in Frontend, Smart Contract development, DEX, DAPP, NFT, setting up ICOs as well as developing wallet applications and trading bots. I am expertise in React, Vue, Web3.js, Ether.js, Solidity, Rust and Blockchain fundamentals, I can help you with most of the crypto-related tasks and bot-related tasks, setting up blockchain explorer, generating your own currency, or writing smart contracts for token and swap development.
If you are currently hiring a Blockchain Developer or Full Stack Developer, please DM me. my Github id is hiccup1991
17:54
The merkle strategy is actually worse than that, because we are storing all the oracle results ever in the oracle merkle tree.
17:57
Using the RSA accumulator this way has the advantage that we can assign those 10k oracles numbers 0-9,999 as they are being created.
So this means that the prime numbers we could need to multiply with, they are going to be one of the 10k smallest prime numbers.
These are the smallest 10k prime numbers, they are the cheapest to store, and they are the cheapest to do finite field operations with.
18:00
I wonder what it would mean if we did this for the 5k most recently accessed accounts?
So if you recently received money in your account, you can quickly spend it again, without needing to pay for an expensive merkle proof. Because your account is in the RSA accumulator cache, you can use a small RSA proof.
18:18
Zack
oh, I guess doing it for accounts doesn't make sense, because they are mutible
18:19
we could probably store some finalized contracts this way though.
That way people can withdraw their winnings without needing a long merkle proof for the contract.
18:21
what makes maintaining one of these RSA accumulators cheap is that the full node should never need to remove something.
If we wanted to mutate an account, they would need to re-feed all the data into the accumulator, and calculate a new root.
20:03
O K
Coin has been through many evolutions in mining. Can easily design asics for this coin, maybe you can take it to the next level.
21:14
Zack
I think upgrading the merkle tree to support rsa collapsing wouldnt be too hard.
Each node needs to know a prime number for all the up to 1024 descendents it can have after 10 layers.
And that might seem like a lot, but each number is only being stored 10 times.
When you update the merkle tree, you only update these lists of numbers for nodes that you would be updating anyway.
If a node is the Nth descendent from the merkle root. And N mod 10 is 0, then you need to store an rsa root if you update that node.
The rsa root is as expensive as multiplying all the numbers we store with that node. Up to 1024.
By adding these few numbers to the merkle tree, it becomes possible to produce the rsa verkle proofs.
And if we store the rsa root with each block, it is possible to verify them too.
Which would let us collapse proofs from log_2 to log_10.
So if there are 8 billion accounts. Instead of 512*33 bytes per proof, there would be 512*3.3 bytes per proof.
Collapsing these proofs is computationally expensive. But it is also completely parallelizable.
If a miner wants to gossip their new block quickly, they could either have proof collapsing hardware, or send the block to a specialist who will collapse the proofs and rebroadcast.
21:23
I think this does not work, because it breaks the stateless full node plan.
How can we calculate the next state root without reading the hard drive, if we need to multiply all those numbers together for the next rsa root?
21:33
Maybe the best strategy is the semi-centralized server that posts all the oracle results to rsa accumulator that it signs and publishes somewhere.
Maybe they can also set up security deposits, where they agree that if their accumulator has a report that disagrees with the oracle on amoveo, or if they fail to include an oracle result, that they lose their bet.
So if we want to bet on a private oracle, we would buy this security deposit for the length of time of the bet, and if the rsa accumulator lies, we are still compensated, and we still arent revealing which oracle we bet on.
21:34
What is nice about semi-centralized rsa is that generating proofs is very cheap for them.
So we can have billions of oracle results in the accumulator, and still generate proofs as cheap as ever.
03:19
Zack
Amoveo works well if you are developing open source software, because the oracle reporters can check if you delivered on your promise.
If you promise to mail a physical item somewhere, how will oracle reporters know if you did or not?
03:19
Z
Idk one user could post “build a shelter for stray dogs in x location with yzetc parameters” and put down some collateral
If the market likes this charity / project / business venture then it could be funded and etx
03:20
Zack
As long as there is some way for oracle reporters to check if the shelter exists
03:20
Z
I dont really understand “oracle reporters” what i could understand is “certified auditing”
03:21
There could be a neutral party that for a fee audits the process
03:21
It could be the community itself maybe? Sleuths and OS info?
03:22
Zack
That neutral party could publish their assessmwnt online. And the oracle reporters could then view that assessment to import the data to the blockchain
03:22
Zack
So you just explain about the neutral party in the oracle question
03:45
Zack
rsa accumulator.
We can store up to around 12 kilobytes in the accumulator for every 64 bytes of extra we keep in the block.
We can prove any data from the accumulator using 64 byte.
We cant delete or edit anything in the accumulator.
03:46
12 kilobytes could be 50k oracle results.
What else could we use it for?
03:53
Živojin Mirić
Neutral does not exist
03:54
Everyone is corrupted, zero trust
Yuliani Susanto invited Yuliani Susanto
08:11
Zack
I think rsa accumulators could make the smart contract language more powerful.
Currently we are using the merkelized abstract syntax tree strategy. So code is broken into pages, and you can call other pages by referencing the merkle hash of the code on that page.
Except for anonymous functions, we need to write the hash of the page on another page to be able to call it. So the pages make a merkle tree pointing to each other.
if your total code is length N, then the stuff that gets written on-chain will need to be at least log2 (N) for the average case.
We could store the pages of code in an rsa accumulator instead of in a merkle tree.
So when you call the contract, you reveal whichever pages you need. And there can be different pages that give different definitions for the same function.
Abstracting on this idea.
The rsa accumulator can be thought of as a kind of multi-signature. Where you are signing a bunch of things at once.
And if you look at just the signature, and any of the parts of what was signed, it is still a valid signature.
08:12
Imagine i used an rsa signature to sign a contract with my bank.
Then the bank could prove to a third party that i had signed any of the clauses in that contract, and the bank doesnt need to reveal anything else about the contract besides that one clause.
10:59
Zack
Maybe governance values should be in an rsa accumulator. Since they almost never change, and it would make the proofs a lot shorter.
And when contracts expire, they could be stored too.
Benjamin Omar invited Benjamin Omar
23:53
The limitation with rsa is that they are append only. You cant edit.
But that is how everything works in immutable languages, and it isnt an issue. I reused a simple trick from immutable programming to enable creating something like a merkle tree, but with much shorter proofs.
00:10
Zack
I guess the storage cost on provers is too high.
Maybe i can adjust some numbers to get it to work.
00:26
Zack
Looks reasonable now
00:26
If this works, it is a really great tool for amoveo. Because our merkle proofs are like 90% of the block space currently.
08:25
Zack
It seems like RSA trees do technically work.
The problem is that for it to be secure, you need to choose a really big group. And that means all the proofs and roots, they are like 256 bytes, much bigger than a hash for a merkle tree.
So an RSA tree is only better than merkle trees if the tree is very big, with many billions of elements, and the RSA tree nodes are very big, with like 2048 or more elements.
02:55
Zack
I looked into the elgamal crypto system over elliptic curves, using pedersen commitments to make an accumulator. It seems like it works.
I put it into test(2) on this page. near the bottom.
https://github.com/zack-bitcoin/homomorphic-tools/blob/master/elgamal.erlSo this is a way to do something like the RSA tree, but with much shorter proofs and commitments. It looks a lot faster to calculate too.
02:58
What is up with the R values? Isn't that a lot of information to store?
Can we just choose 1 every time? and then R tells you how many values you have accumulated over?
03:02
A proof is 3 numbers.
V which is a sum of all the values you have accumulated over.
R, which is usually a random value. but maybe for us it can be the total number of values accumulated over.
and C, which is an elliptic curve point.
V and C are both probably going to be 20 bytes, if we use curve25519 for the elliptic curve math. and R is something like 3 bytes, if we use that RSA verkle tree design I had outlined. So 43 bytes per node, vs 64 with merkle trees.
To make that verkle tree design work, we do need to find a way to prove that a value has not been accumulated over.
06:27
Zack
my accumulator still does not work.
07:47
Zack
I think it works now
07:47
I needed to make a different generator group point for every variable we can store in the accumulator.
07:50
it looks like this is going to work for editing and deleting elements too. we don't have to do the trick from RSA that made the cost 3x.
And we don't have to refresh the nodes like we would have had to in RSA.
08:52
Zack
It seems like this pedersen accumulator tree can get us 95% of the way to perfect, and all the intense math they are doing at ethereum, the KATEs and zkSNARKS and probabilistically sampling over polynomial commitments, I think it is only giving them that last 5%.
If we can at least get the proofs to be smaller than the rest of the transaction, then they wont be the bottleneck any more, and we can move our focus for now.
09:04
For every layer of the pedersen tree, lets go with 1024 elements.
The proof needs 2 elliptic points for every layer of the tree. each point is 20 bytes if we use curve 25519
So if there are 1 trillion users, that is 160 bytes per proof.
Compared with merkle trees, which would use 2560 bytes per proof. Or our current merkle tree implementation, which is around 2x worse than that.
If there are 1 million users, pedersen trees need 80 bytes per proof, and merkle trees need 1280 bytes per proof.
The average transaction in amoveo needs around 2.5 proofs.
Transactions will use like 1/10th as much block space.
After an update like this, it would make sense to use the shorter address format, and look into using the compressed pubkey format when we need to write the pubkey, and look into whether there is a library that makes shorter signatures. Because all these other things will become the bottleneck instead.
Given how long all these other parts of the transaction are, it seems like there is little to gain from making the proofs even shorter.
We can wait for other projects to experiment first.
Jeffrey Dean invited Jeffrey Dean
19:46
Zack
I found a log2(N) algorithm for maintaining and updating individual nodes in the pedersen tree. where n is the number of children per node.
So that breaks the bottleneck on how big each node can be.
It might make sense for a node to have over a million children.
So even if our merkle tree had 1 trillion elements, each proof is only 2 nodes long. ~40 bytes.
Because all the proofs are reusing the first node, we can compress that into a single 20 bytes.
So we can make a multi-proof that is ~20 bytes per fact that we are proving.
19:47
For the first million users, all the proofs are only 1 step long, and we can compress all those proofs into a single 20 bytes.
19:52
I wonder if we can have an expandable pedersen accumulator. So we would deterministically create a new group generator for every new thing that we want to store.
Then we would have constant sized proofs per block.
23:24
Toby Ganger
I think the best use case for VEO is pontificating about the future use of VEO even though our community is basically non-existent and it’s virtually impossible for newcomers to mine…maybe we can have an algorithm alled Proof of Pontification? It’s like Proof of Work but for autists
23:25
Zack
Shorter proofs is very important for stateless blockchains.
Amoveo was the first stateless blockchain, and it is important that we stay up to date with this tech.
03:12
yeah my multitx combine worked xD
06:00
john blair
I'm a idiot but it seems like your doing good work zack, keep it up. I hope you have someone to help you with the tasks your set out to do.
Jesse Nea invited Jesse Nea
11:12
mr x
===> Dependency failure: source for light_node does not contain a recognizable project and can not be built
11:13
when i try make prod-restart
13:20
Zack
I think you can ignore that error. or does it not build? I had removed some files I had thought were unnecessary from the light node.
13:23
mr x
Makefile:138: recipe for target 'build' failed
13:24
Zack
Ok, ill get to my computer and fix it in a few hours. Thanks for noticing.
17:27
Zack
it should be fixed now.
But you need to do a ./rebar3 upgrade to make sure you have the new version of the light node before you build it again.
19:50
Zack
https://github.com/zack-bitcoin/amoveo-docs/blob/master/blog_posts/optimistic_rsa.mdI learned more about the kind of verkle tree they are making for Amoveo.
I decided to reflect on these verkle proofs from the optimistic setting, like how optimistic rollup works.
Under optimistic proofs, it seems like a much more efficient database is available to us. I call it the optimistic RSA accumulator.
In this blog post I compare it to the verkle tree idea.
22:48
Zack
My idea didn't work. you can't combine partially made rsa proofs. I guess im going to be learning bullet proofs.
22:57
mr x
ok synced now but cannot connect light node to it hmm
22:57
Zack
I probably forgot to add a page of the light node to the full node's whitelist
22:59
im having trouble reproducing this error
23:01
Zack
oh, I got it.
When I run the light node from my backup server, it isn't working
23:02
oh, I think some of the links are just broken
23:03
looks like I didn't push some changes to the full node about how the links moved around
23:05
you need to update the full node to support this.
23:35
but get server rejected when sending
23:37
did you do sync_mode:normal(). and keys:unlock(""). on your full node?
23:37
the full node doesn't accept txs when it is in fast-sync mode
23:37
maybe unlocking your key isn't necessary
23:38
now everything works !
07:50
Even with all the math being done in plain erlang, nothing compiled. Even using this slow elliptic curve that hasn't at all been optimized for speed. it seems to take 13 seconds to make and verify a 256 member vector commitment.
Maybe ethereum is optimizing their vectors too short. Maybe 2048 would be better.
08:05
verifying is almost instant
08:11
23> pedersen:test(3).
proving 256 values.
verifying proof of length 17
proving took 14.47422
verifying took 0.39718
success
08:11
14.5 seconds to prove, 0.4 to verify