sipa changed the topic of #bitcoin-wizards to: This channel is for discussing theoretical ideas with regard to cryptocurrencies, not about short-term Bitcoin development | http://bitcoin.ninja/ | This channel is logged. | For logs and more information, visit http://bitcoin.ninja
oniijin has quit []
<yanmaani>
Also, I am unsure whether negative reputation would be a very good incentive. On *ahem* traditional markets, a vendor bond of $200, a fee of 4%, a yearly gross revenue of $1 million, and an average order time of 1 week may be common.
<sarahc0nn0r>
yanmaani the exit scam prevention is discussed in the paper in detail and is pretty robust in the sense that it’s uneconomical to do one as long as users don’t buy from merchants with negative reputation.
<yanmaani>
This means on average there will be $19k in flight at any given time (1m / 52)
<sarahc0nn0r>
yanmaani, what is the issue in your view with review text being public? It's pseudo-anonymous like transactions in Bitcoin and the paper discusses in detail why this is safe for the average user.
<yanmaani>
On the first $100k will be collected $4k in fees for example. With vendor bond of 200 and additional funds needed to put it at $19k, the total volume will have to reach (19230-200)/0.04 = $475k
<yanmaani>
So the first half a year the vendor will be at negative rep.
<yanmaani>
With traditional markets this is no problem, as they have escrow system and prohibit FE. But here? I am unsure
<yanmaani>
sarahc0nn0r: Well they will. The vendor bond will have to be extreme, since all your transactions are going to be FE
<sarahc0nn0r>
yanmaani, it's a good point that merchants will generally need to "warm up" to being able to have a large "in-flight" capacity without going negative. That being said, does this seem like a major problem to you?
<yanmaani>
Well, the result will either be that nobody uses it or that people trade with negative rep.
<sarahc0nn0r>
Yeah, it seems like the reputable merchants will get away with negative rep initially since people know who they are by their public key.
<sarahc0nn0r>
That being said, it seems all of this is moot to the extent that traditional marketplaces are constantly going down whereas the Ultranet will retain your reputation forever.
<yanmaani>
I mean a normal escrow system would just solve this. This is how normal markets do it.
<sarahc0nn0r>
As in if you're willing to warm up on the Ultranet, you'll never have to deal with another market again.
<yanmaani>
"Traditional marketplaces" have already attempted to deal with this by using 2-of-3 multisig
<yanmaani>
well the sellers go where the buyers are, and the buyers are where the sellers are. You still have netowrk effects so it does not seem so convincing
<yanmaani>
If everything is public, what is the upside over just using regular forum posts?
<kanzure>
yanmaani: is there any insightful progress here on reputation systems or is this just an altcoin that needs to be kicked from the channel?
<yanmaani>
kanzure: Read the paper, I am not going to suggest anyone be kicked.
<kanzure>
whitepaper looks like mostly filler. on that basis i'm inclined to kick.
<sarahc0nn0r>
kanzure, I think if you read the section on "How Merchant Reputation Works" it will answer your question.
<yanmaani>
It is verbose, I suppose?
<yanmaani>
sarahc0nn0r: how does the 20% discount work?
<sarahc0nn0r>
yanmaani, it's a bit of a "bootstrapping" feature. Basically when you use the app it shows you "price merchant set"+20%, but then when you actually create an account it shows you the actual prices (regardless of whether or not you enter the public key of your referrer)
<sarahc0nn0r>
You can see it if you check out the demo on the site: http://ultranet.one
captjakk has quit [Remote host closed the connection]
<yanmaani>
Yeah but isn't this just dishonest? Why would I use someone else's reflink when I could just use my own?
captjakk has joined #bitcoin-wizards
<yanmaani>
it's no better than random websites which show you prices and then multiply them by rand(1.2, 1.8) and show "50% off"
<sarahc0nn0r>
You can. You can also just not enter one, which has the same effect (it just says "ok, you're going to get the 20% discount regardless")
<yanmaani>
Is there any way to pay the inflated price?
<sarahc0nn0r>
That's why I said it's a bit of a bootstrapping feature. It just gives people who tell their friends about it some motivation to get them to use it. Namely "hey if you enter my key you can get 20% off."
<sarahc0nn0r>
Without needing a centralized entity to enforce anything.
<yanmaani>
By "bootstrapping feature," do you mean "lie"?
<sarahc0nn0r>
No, it's basically impossible to ever pay the inflated price.
<yanmaani>
Is it theoretically possible to do it? Or is it just for show?
<sarahc0nn0r>
It's not possible to pay the inflated price.
<sarahc0nn0r>
If you don't enter a referrer key, it still gives you the discount.
<sarahc0nn0r>
It's just an implementation detail honestly; I wouldn't fixate on it.
<yanmaani>
Isn't it just a lie then? If I tell my friends that they get 20% off if they use my key, this is not true.
<yanmaani>
I mean, I hate to be that guy, but the dictionary definition of a lie is, "a false statement made with deliberate intent to deceive"
<sarahc0nn0r>
Well, it's true that entering your key will give them the discount. It's also true that they could also get it by pushing through the UI without a key. But again all of this is just implementation detail not actual technological improvement, which is what I'm seeking feedback on.
<yanmaani>
Their incentive to use my key is the 20% discount, which supposedly compensates for the 1% fee they pay back to me. But if the discount does not exist, then they are just lying to me - they have no good reason to give me the money
captjakk has quit [Ping timeout: 258 seconds]
<yanmaani>
My honest feedback is that this does not solve a problem people have. On traditional marketplaces, there is a good amount of safety for customers
<yanmaani>
As well as vendors
<yanmaani>
Vendors can set it to automatically withdraw, and it's possible to use 2-of-3 multisig for escrow
<sarahc0nn0r>
I appreciate that. What do you think of the ciriticsm that traditional marketplaces are constantly being shut down?
<yanmaani>
If they would use 2-of-3 multisig it would not be a problem. It would also not be a problem if they did legitimate business.
<yanmaani>
from a purely practical POV, the UX here is already dreadful (download TBB, buy Bitcoins)
<yanmaani>
making this even worse (download blockchain, allocate 25 gb, buy bitcoins, buy ultra) will not be good for adoption
<yanmaani>
The incentives just aren't in place. Why not just use a regular forum?
<yanmaani>
The advantages of standard markets over random phpBB forums is that you get escrow, anonymous reviews, and curation
<yanmaani>
if you don't get any of this but still pay 4% fees to enrich random holders of a cryptocurrency
<yanmaani>
why bother?
albertlast1 has joined #bitcoin-wizards
<sarahc0nn0r>
yanmaani the core value proposition is that it can’t be shut down and the data, in particular the reputation data, will therefore never be lost. Along these lines, it seems like even if the UI is equal (IMO it’s quite a bit better but say it’s equal for argument’s sake), it seems like eventually, as other markets get shut down, the
<sarahc0nn0r>
community will slowly aggregate onto this platform. But I’m honestly curious what you think about this dynamic. Thanks for the feedback overall btw.
<yanmaani>
also where does it explain how it is a good idea to give linkable reviews? Letting a vendor who has a buyer's personal info see the buyer's other purchases is a poor value proposition
<yanmaani>
fundamentally, a marketplace with buyers but no sellers will attract sellers to a greater degree than a marketplace with sellers but no buyers; sellers are willing to jump through more hoops as they do it on an industrial scale
<yanmaani>
(the reputational data decays, so this seems rather flawed)
meshcollider has quit [Ping timeout: 268 seconds]
<yanmaani>
(If you'd make it so that burnt reputation is a token that can be traded and does not decay, then you might have something.)
<sarahc0nn0r>
Sure, but the argument is that the buyers *and* sellers will aggregate as traditional markets get shut down over time.
<yanmaani>
so, sellers use TAILS and complicated money laundering schemes; buyers send straight from exchange to market using onion.to
<sarahc0nn0r>
Interesting. Could you elaborate on your thoughts on the tradable reputation component?
<yanmaani>
yeah but the UX is not so good. There are no hard guarantees for the buyer not to get ripped off.
TheoStorm has joined #bitcoin-wizards
<yanmaani>
Traditional marketplaces put a lot of effort into this
<yanmaani>
Banning FE, ensuring escrow is always used, vendor bonds, active policing, having support chat
<yanmaani>
here, this is replaced with "dude game theory lmao". I am not convinced
<yanmaani>
sarahc0nn0r: yeah so if a vendor has reputation that he had to spend $1000 in fees to get
<yanmaani>
this means he can safely exit scam $1000 worth of orders while still having a positive rep
<sarahc0nn0r>
So it seems like you're saying that if the traditional marketplaces are shut down, leaving no alternative, they will use the Ultranet but then go back to the first traditional marketplace that comes online, even if the UI is equivalent?
<yanmaani>
so he has the choice between exit scam (hurting buyer confidence) and no exit scam (costing him $1000)
<sarahc0nn0r>
Just trying to make sure I understand
<yanmaani>
thus he is more or less forced to exit scam
<yanmaani>
whereas if he could sell his rep, he'd have two choices; exit scam or sale
<yanmaani>
the latter would earn him a strictly greater amount of money, be more honorable, and attract less scrutiny
<yanmaani>
thus it should dominate
<yanmaani>
sarahc0nn0r: Well, except for the part where they will never use the Ultranet.
<yanmaani>
There has not ever been a case when there are zero marketplaces out there.
<yanmaani>
If you want something interesting, look at the Nash equilibrium market ideas. And Paul Sztorc's suggestions for making REP tradeable.
<yanmaani>
so the UI is not equivalent; users want good assurances they are dealing with serious professional people
<yanmaani>
in retail markets like ebay/paypal, we see this clearly: payment processors and merchants bend over backwards to satisfy customers
<kanzure>
has hanson or sztorc ever put together a functioning WoT reputation market or prediction market?
<yanmaani>
since "please send your money here and you might get something if you are lucky" will not be a very good business proposal
<yanmaani>
kanzure: sztorc has put together some code
<sarahc0nn0r>
yanmaani it seems we disagree about the state of traditional marketplaces. In my view there have been many instances in the past few years when the alternatives were virtually zero. But I understand this is not something that's easy to come to consensus on, and I appreciate your side of the argument.
<yanmaani>
hanson has made one I think but play money only
TheoStorm has quit [Quit: Leaving]
<yanmaani>
sarahc0nn0r: Look at India marketplace I think it was called; they had 2-of-3 multisig. They still exit scammed with the like $5000 of custodial money. But in theory that system works.
<sarahc0nn0r>
Taking a look now. I'll dig into the India Marketplace as well and see what they were doing. Other than the "warm up" period, though, was there anything you saw that would be problematic about the current reputation scheme though?
<sarahc0nn0r>
Appreciate the feedback overall yanmaani
shush has quit [Remote host closed the connection]
fiatjaf has quit [Ping timeout: 256 seconds]
fiatjaf has joined #bitcoin-wizards
Guyver2_ has joined #bitcoin-wizards
Guyver2 has quit [Ping timeout: 240 seconds]
shush has joined #bitcoin-wizards
Guyver2_ is now known as Guyver2
nick_freeman has joined #bitcoin-wizards
nick_freeman has quit [Ping timeout: 256 seconds]
DeanWeen has quit [Ping timeout: 240 seconds]
aupiff has quit [Ping timeout: 256 seconds]
JackSparrow has quit []
dgenr8 has joined #bitcoin-wizards
aupiff has joined #bitcoin-wizards
michal_kubenka has joined #bitcoin-wizards
shush has quit [Remote host closed the connection]
shush has joined #bitcoin-wizards
nick_freeman has joined #bitcoin-wizards
nick_fre_ has joined #bitcoin-wizards
francisco_______ has quit [Quit: Connection closed for inactivity]
shush has quit [Ping timeout: 240 seconds]
shush has joined #bitcoin-wizards
nick_freeman has quit [Ping timeout: 256 seconds]
shush has quit [Remote host closed the connection]
shush has joined #bitcoin-wizards
shush has quit [Read error: Connection reset by peer]
shush has joined #bitcoin-wizards
aupiff has quit [Ping timeout: 256 seconds]
captjakk has joined #bitcoin-wizards
captjakk has quit [Remote host closed the connection]
captjakk has joined #bitcoin-wizards
captjakk has quit [Ping timeout: 256 seconds]
DeanWeen has joined #bitcoin-wizards
mauz555 has joined #bitcoin-wizards
michal_kubenka has quit []
rjected has quit []
davispuh has joined #bitcoin-wizards
peterrizzo has joined #bitcoin-wizards
bpalmer1 has joined #bitcoin-wizards
real_or_random has quit [Ping timeout: 240 seconds]
mdunnio has quit [Remote host closed the connection]
real_or_random has joined #bitcoin-wizards
<gleb>
I found the work "BitML: A Calculus for Bitcoin Smart Contracts" which I don't recall being discussed, like, at all. It seems to be fully compatible with current Bitcoin Script. I'm wondering if that's something to compare with miniscript and whatnot digi_james
davispuh has quit [Ping timeout: 255 seconds]
davispuh has joined #bitcoin-wizards
<digi_james>
Hi Gleb. I think BitML is a very elegant approach to a contract specification language for Bitcoin, because it is basically a process algebraic description of the underlying transactions enforcing the contract. The compilation to Bitcoin txs is therefore quite straightforward. However, I think there is an opportunity for different compiler strategies which compile to txns which exhibit equivalent contract
<digi_james>
settlement properties, but optimized for privacy or fees.
davispuh has quit [Client Quit]
davispuh has joined #bitcoin-wizards
Guyver2 has quit [Quit: Going offline, see ya! (www.adiirc.com)]
<sipa>
gleb: afaik bitml only compiles to scripts (and across multiple transactions); miniscript is more about reasoning about scripts that potentially come from multiple sources, and includes also signing logic and determining the policies involved
slivera has joined #bitcoin-wizards
zmnscpxj has joined #bitcoin-wizards
Kiminuo has quit [Read error: Connection reset by peer]
Kiminuo has joined #bitcoin-wizards
<digi_james>
Hm, I suppose BitML is a contracting abstraction vs Miniscript, which is a script abstraction.
orange_ has joined #bitcoin-wizards
<digi_james>
Miniscript would allow the BitML compiler to compile to fewer txns, if implemented.
gleb1 has joined #bitcoin-wizards
gleb has quit [Ping timeout: 255 seconds]
shush has quit [Remote host closed the connection]
<roasbeef>
gleb1: in a sense, it's more powerful than miniscript, as it allow expression of _transaction level_ templating vs just in the script itself, so I could write out some multi-stage/transaction off-chain protocol using this (output is set of transactions to be signed) and then even possilby generate a state machine from this as well