Taking a step backwards, we have a formula: `privateKey*GeneratorPoint=publicKey` but for this to even be useful, we know if we're given a publicKey there's must be no real way to figure out what the privateKey is. If there was a relationship like the publicKey went "higher" when the privateKey went "higher" we would be very quickly able to bisect what the privateKey was.

So maybe it helps to understand there absolutely must no be this relationship, and then dig into how ECC works to do that. I think the real trick to make all this magic work, is that everything is mod'd by p ( 0xfffffffffffffffffffffffffffffffffffffffffffffffffffffffefffffc2fn ). Which I think if you work with "real numbers" you'd have the same effect as it basically causes the numbers to wrap around when they get too big.

I haven't had any involvement with bustabit in years, so just speaking from a purely personal point of view:



Years ago, I did an experiment where I used a site (leakedsource) to try see how secure accounts were. Of people with a balance and set email address (so I could search easier, which albeit wasn't public), I was able to guess about ~30% of all accounts passwords (after multiple attempts) before mandatory passwords were enforced (and the site displayed a huge amount of warnings about picking secure and unique passwords).

The simple fact is if you let people pick their own password, a large fraction of people will pick extremely insecure passwords, get hacked, and then constantly bitch about how they got hacked. Or you can force people to use secure passwords, and listen to people constantly bitch about their rights to reuse their already leaked and insecure password.

But honestly, if you are so moved about the desire to have poor security -- look at your bitcoin wallet --- they are forcing a hard to remember seed on you.  I suggest you instead find a brainwallet, so you can have the freedom to pick your own seed from the password you normally use. It's also pretty convenient to just get hacked directly from your bitcoin wallet without needing to involve third parties at all.

Because as a rule, the people it inconveniences are the people who don't normally use secure, unique passwords.  Those are the people whose passwords are probably already leaked (see: https://haveibeenpwned.com/ ) and then wonder how someone else knows their password. Auto-generated passwords have no real impact on people who already have good security practices (i.e. use a password manager with unique password). In the old days you could "hack" the client by opening devtools and changing the password to what ever you wanted and submit the form. Not sure if that still works, or really why you'd want to. Just use a password manager like the cool kids


Yeah, there's a wagering threshold to stop spam, I believe.

It's just a joke. All it does, is literally make the games green. It's a purely cosmetic thing Daniel recently added as a joke for the people who love green games

Yikes, that would be mighty disappointing if true. It's explicitly illegal in a lot of places (e.g. Nevada); and for pretty good reason.


One of the things I really liked about the early days of bitcoin gambling was even in the absence of regulation, the major players were reputation-driven and there was virtually no bullshit and it was clean/honest/transparent. You could even give them your email without fear of being spammed by promo shit, etc.

I guess it's a bit like that batman quote: "You either die a hero, or you live long enough to see yourself become the villain."

Sure, I'm not accusing roobet of anything. Just (openly) asking the question to get an official response  


Well, it depends on what country you are talking about. But it's quite illegal in many countries I'm aware of. But more importantly, it would be extremely unethical. Talk to any of the reputable casinos, and this is something that absolutely would not be done.

Someone told me that roobet gives streamers and promoters free pretend money, so they stream/promote roobet with fake numbers .. and as payment  for the promotion they can only cash out a fraction of their "winnings".

I would normally not believe it, but the person who told me is pretty trustworthy. Is this really happening?

The maths is reasonably simple, I think.

Lets say in this example bankroll is 4000 BTC. The first calculation is the easiest, Daniel has a commissions rate of "bankroll / 10000btc".  So in this example, the commission he charges is 40%. This means when ever the bankroll profit increases past its previous all-time-high ... Daniel charges that 40% and updates the new all time high profit.

Now let's say you want to invest 1000 BTC ... the new bankroll will simply be the old bankroll + how much you invested. Thus it'll be 5000 BTC. And now Daniel will be charging 50% commission (see formula above).

But the important number is your stake: You put in 1000 BTC the 5000 BTC  [I'm ignoring the actual dilution fee to keep it simple], so your stake will be 20%. This means you actually own 20% of the bankroll. This is the important number, because it only changes when other people invest or divest. If people win money from the bankroll, your stake stays the same. It's just the bankroll has decreased. And likewise when the bankroll increases, your % ownership is the same ... just the actual bankroll is bigger.

So let's imagine that the bankroll is 0.5 BTC away from it's previous ATH profit -- and someone gambles and loses 1 BTC. Daniel will charge 0.25 BTC in commission (50% of the 0.5 BTC it exceeded previous profit ATH). Thus the bankroll will increase by 0.75 BTC. You own 20% of that bankroll, so you have effectively made 0.15 BTC.


People on average will lose 1% of what they wager (to the bankroll).  So if you can guestimate how much people will be wagering, you can figure out what the expected returns will look like.  

By any chance, does the winning strategy involve investing in the bankroll?

I think the question is so underspecified its hard to tell what they're asking exactly. I think you're answering it from a performance-point of view. While I've only done stuff on top of bitcoin, where I run into problems that bitcoin-core isn't designed-for or flexible enough for my use cases. Probably the most classic example (but far from only) would be not able to lookup transactions that are from/to a particular address. So when ever I've done my stuff, I've tried to optimize for flexibility/developer experience -- but never had to even consider performance. For my usage, even if it took an additional 10 seconds (which it doesn't) to process a block, it would be no big deal at all.



I'm not disagreeing with you, but it almost feels like you've reasoned backwards about it. If you model the "blockchain" in a very natural way (i.e. blocks have transactions and parent blocks). (transactions have inputs and outputs) with the right indexes and what not, you can do all the operations you need in an efficient (i.e. sublinear) time. i.e. most operations you talk about is logically "does this output exist in this chain" & then "and has it not been spent in this chain".

While the UTXO feels very "unnatural" in comparison, as you're basically manually maintaining a index for a very specific question, and that without some sort of "structural sharing set"  -- your utxo is not going to be efficiently support re-orgs.  

Of course it's the right choice for a high performance full-node. But if you were building something that needed more flexibility (e.g. wallet node, block explorer) or trying to build something as correct as possible with the least amount of code it might not even make sense keeping your own materalized utxo set.

---
A bit of a tangent, but bitcoin-core's real problem is it doesn't by come with a jump started utxo set. It's kind of insane that everyone is supposed to build it themselves when if each release came with a new utxo set hash hardcoded in, it'd be easy for dozens of people to verify it and then people would actually be able to use/recommend bitcoin-core instead of electrum or w/e

A week ago DrDoofenshmirtz released an open source beta version of a "chaum bank" called moneypot. The release has flown under the radar, but I think it's a pretty significant development.


While it's not designed for it, I think an interesting aspect is that it accidentally totally and completely obsoletes bitcoin mixers. It has all the positive aspects of a bitcoin mixer (ease of use, non-repudiation) while offer huge advantages:
* cheaper
* there's no need to trust the mixer to not record/share/save/sell the link between inputs and outputs. On moneypot the custodian literally doesn't know themselves
* it's a useful wallet in its own right
* less tainted pool of coins
* all code is opensource, even the custodian


As far as I can tell, moneypot.com is strictly better than any bitcoin mixers. I don't see a single advantage. Hence: bitcoin mixers are obselete, it offers zero advantages over a privacy wallet like moneypot and a lot of disadvantages. Anyone want to change my mind?



Disclaimer: I am emotionally invested in the project. I however have no financial interest in it, nor have any involvement with the operations or running the custodian.

I'm going to throw postgres out there.

I've used it for a lot of bitcoiny things, including a wallet that I never finished -- and I've always been extremely happy with postgres. Having a real database makes for an amazing developer experience, especially debugging. You don't need to create an API for everything, you just query the data. And a new access pattern often just means adding a new index. It has great support and documentation for concurrency and bona fide transactions with meaningful isolation levels.

It easily scales to the size of bitcoin (e.g. storing all transactions, while indexing the utxo, etc) and in general I think is a lot more performant than KV databases. Maybe not for some micro benchmark, but postgres makes it feasible to actually make sure every action you're doing can run in sub-linear time. While the same is technically true for lower-level datastores, postgres actually makes it actually realistic.


I think the only real downside to postgres is deploying/packaging it for end users. That's nightmare material; and most likely a show-stopper for most use cases. Other than that, it's pretty cool though 

In theory, it looks quite doable. The only two things you can do with your blinded coins:

A) Make a bitcoin transaction ( https://github.com/moneypot/moneypot-lib/blob/master/src/hookout.ts )
B) Make a lightning transaction ( https://github.com/moneypot/moneypot-lib/blob/master/src/lightning-payment.ts )

which then you give to a custodian, who processes it. So it would be pretty easy to make a "soft fork" style custodian, that simply added additional restrictions (e.g. only pay lightning invoices that were to do with an internal payment). Or you could go all-out and make a special transaction type, and modify the wallet, etc.

---

But on a pragmatic note, I don't think it would make much sense. The complexity involved is *insane*. And trying to do any customer-support would be a lolsfest, (compared to now, just doing some db lookups).


The only case I can think it could maybe make sense, would be as a "legal hack". Like I'm imagining that if you were in a regulated industry who is required record keeping (and KYC?) you could technically comply with those laws, while also being able to offer a huge amount of privacy.  But you're always running up against a stupid amount of complexity. So you really need an application where privacy is genuinely extremely important and its a real selling point

Great seeing this finally released, I know it's been years in the making. I took it through a spin, and it's a bit rough (like states not automatically updating or showing that they're updating), but it works as advertised and is quite a big break through imo

It's absolutely insane to me that google cares so little about advertising phishing/malware/fraud. It's pretty much willful negligence.

They only take action after several days and lots of complaints -- only to allow the exact same thing to happen again on a different domain in an endless cycle.

Honestly hope someone takes google to court over it. Even if you can't win at court, I'm sure google would find it rather embarrassing to try justify why they have not duty to do even the most basic diligence when taking money to show someone an ad for a product that the person is clearly not trying to get to.


This is also why I strongly recommend that everyone use an adblocker, and make sure all your friends and family are doing so too



My suggestion for bustabit: Try detect adblockers, and if the person doesn't have one -- encourage them to download one!

If you did that, you'd get:
for bustabit, and:

for bustadice.


However considering how much bitcoin has gone up recently, I don't think people will continue wagering nearly the same amount (when measured in btc). FWIW earlier today in the chat Daniel said he guesses yearly p.a. yields are going to be around 13%. I'm not sure how he calculated it, but it's reasonably close to my first estimate.


But seems like no matter how you cut it, bustadice is offering better yields than bustabit.

Interesting. Let me try roughly calculate it:


It's not a perfect formula, but i think it's correct enough (?!).

So trying for Bustabit (i'm just guestimating 400 btc a day volume, not sure how accurate that is):

and trying for Bustadice (i'm just guestimating 100 btc a day volume, not sure how accurate that is):

I think the bankroll hasn't changed too much. It's just the bustabit has used the terminology "bankroll" to mean "onsite + offsite", and on the 15th of December everyones offsite was set to 0 (and they were given dilution fee credits for that amount). So the bankroll dropped a lot, but the amount "onsite" stayed the same.

It'd be nice if the /stats page showed the commission directly, but currently the bankroll is 4956 BTC, which means the commission is 49.56% (or slightly less than before the change). On bustadice, the bankroll is considerably smaller (1977 BTC) which means commissions massively dropped to 19.77%.

The first (organic) result is legitimate, it's just people buy ads to promote malware/phishing/etc and google puts the ad results above the organic results. By using an adblocker, it removes the ads and leaves only the organic results which are legitimate.

So yeah, install an adblocker and encourage non-tech family/friends to do so too. The amount of actively harmful ads google runs is obscene. And google takes so long to take any action (days or weeks), that it's often more practical to encourage people to click the phishing links to blow their advertising budget.

To be honest, I'm not sure. It is my understanding that he pin is a shared secret, however it previously was not required to initiate a transfer, so I very well could be wrong. Before when they blacklisted the wallet, they allowed the option to enter a pin -- so something has obviously changed. I have backed up the pin and never shared it, so assuming it's a shared secret it could be pretty useful in establishing that I am in fact the real owner.  But then again, if they are now prompting for a pin it's quite likely that people could have brute forced it, the key space is extremely small.


I just tried, and I was able to login and using the correct pin. However it's impossible to even attempt to use the 2fa:



So like I said, no matter how they spin it -- they are screwing the real owner of the wallet.



Edit: I was able to login using the incorrect pin. So I assume the pin is not in fact a shared secret, and just used for local encryption. They probably can't use the pin to verify me

Please do!