The way I know is that I looked at the source code and built my own executable from that source code.

If you have the time and capability, you could do the same.

Alternatively, you can move your coins into a new wallet, then give it the old one. So long as you don't receive coins at any of the old addresses in the future that will be safe even if the wallet is trying to steal from you.

Also bear in mind that even if you don't "give" an altcoin program your BTC wallet, as soon as you run any program on your computer it can grab a copy of all your wallets and send them to the author of the program. So there's no extra trust needed to run this particular program.

Finally, here's a little more proof.

See this address:
https://blockchain.info/address/1MJtFXUBrvMB3yVkVFuTsfKXjEnyxNx7E4

It currently has 41.77661269 BTC on it.

Converting that address into a CLAM address (details):


Check the CLAM address:

http://khashier.com:2750/address/xUcX9QJf8dqD6MbcqjY8JqcJxM6DNd6JL2

You'll see it has been staking for 21 days and has grown from the free 4.6 CLAMs it was given to 5.53249215 CLAM today.

The CLAM wallet could have stolen the 41.7 BTC at any time in the last 3 weeks, and hasn't.

I only ever gambled BTC (27 of them) and DOGE (very little) on your site.

I didn't even know you had a USD option.

There's no "line", just a "difficulty".

Most of the time the difficulty is at its minimum, meaning that every 1 CLAM-day output has a 1-in-a-million chance of staking each second.

Currently the difficulty is 20 times higher than the minimum, meaning that each output is 20 times less likely to stake in any given time period.

So if you're used to staking 20 times per day, now you'll be getting about 1 block per day instead, and it will feel as if you are at the "back of the line".

I usually get 5 or 6 blocks per hour, but overnight I went 5 hours without staking a single block!

The sjess situation has been discussed too much already.

Dean has said that he isn't willing to discuss it any more, and that his decision to refuse to pay him out after changing the rules will stand.

I took a look at the investments page and saw:

You can view the investment cold wallet which contains at least 90% of investment funds here https://blockchain.info/address/1FhxjhFb1YudGC1vtV4FuLpQc2uHzq3XFd

Just the invested Bitcoin amount comes to 677.77159191 and the above address holds 607.39073354 BTC.

That's 89.6%, which isn't at least 90%. If you take the DOGE and other invested coins into account then it's even less.

Maybe that "at least 90%" text should be changed too. When dealing with other people's coins it's best to be as transparent and accurate as possible. Investors spook easily.

That's true.

All I would say is if you're too lazy or too dumb to understand how provably fair gambling works, please don't make comments about it.

Leave that to the people who understand it.

It always seemed to me that accepting 0-conf deposits was asking for trouble.

The difference between 0 and 1 confirmation is much more significant than between 1 and 2.

Just-Dice ran for a year, and in that time we never once saw any double-spend attacks.

We had a few deposits that never confirmed, but since we never credited deposits until they were confirmed, that wasn't a problem.

The reason for allowing players to withdraw with less than 6 confirmations on their deposit is that they like it that way, and the risk of a double-spend seems to be pretty small.

You can't *know* that they won't steal your coins. You can trust them, based on their behaviour so far, etc. Basically it comes down to trust.

Weigh up how likely you think the site is to go rogue against how much you expect to make if they don't, and work out how much you're prepared to risk given those judgements.

I think you're wrong.

Most dice sites are provably fair, which means you can check that the rolls are fairly generated.

You need to do a little work like verifying hashes, picking random seeds and such like, but if you are willing to do that you can be sure the site isn't cheating you.

For the record, I got my withdrawal at "2014-07-29 23:59:49" according to blockchain.info.

I have tried the site a few times. It looks a lot like when you buy a new laptop, before you format the hard drive and install an operating system on it. I guess it's a lot like Windows 8. It looks weird, bright colours, crashes a lot, runs slow, doesn't really work.

It would be good if there was an option to turn off all the weird graphic effects that don't work and just play dice.

I saw this album of screenshots posted in the PRCDice chat too:

http://imgur.com/a/2h1A6

How many things are wrong with this picture?



1. it says I can claim the random bonus NOW but then tells me I can't

2. it tells me a frequency, but gives it in seconds. frequencies are in "per second", not in seconds

3. it tells me a minimum frequency that I can claim which is dumb. there should be a maximum frequency, not a minimum one

4. if I can only claim every that many seconds, that's around every 68 years!

>>> (2**31-1) / 3600.0 / 24 / 365.25
68.04965038532714

Anything else?

I also noticed there's no "client seed" involved in the game. The site decides what number you're going to roll and you have no way to affect it. The site proves that it decides before you roll, but if it sees you are playing "over 95" say, it can decide to give you a lot of numbers over 95 and there's nothing you can do to change it.

That's why every legitimate dice site lets the player specify part of the seed.

I recommend against playing on any site that doesn't let you specify a client seed.

Hey Mike, you quoted my post but didn't type anything...

Mike, don't be a dummy. If you allow a maximum profit per bet of 10 BTC when you only have 20 BTC you will go bankrupt in no time.

Your maximum profit per bet should be no bigger a percentage of your bankroll than the house edge percentage. And even that high is dangerous.

50% is just crazy.

Yes, that could be an issue. But I find that for the most part no more than 50% of my CLAM outputs are immature at any point in time, and usually it's much less, and also the exchange could make it clear that in extreme circumstances coins won't be immediately available for withdrawal. It's comparable to trying to withdraw coins from cold storage - you expect to have to wait while cold stored coins are moved back online.

I guess security could be an issue. In order to stake, you need to have your coins online and effectively not protected by encryption. I know there's an "unlock for staking only" option, but that's not really any protection at all against a determined hacker.

I expect the exchange is staking them.

In fact it wouldn't be hard to find out if you bought CLAMs from an exchange by tracing them back on the blockchain and see if they were staking before you withdrew them.

Edit: Well, OK, it looks to me very much like the exchange *isn't* staking them:



A 138 CLAM output with 22 days of age should stake very easily:

>>> 2**20 / (138.16 * 22)
345

ie. after about 345 seconds. It's inconceivable that it has been trying to stake for 22 days and not managed it.

On the one hand that does seem like a waste, but on the other hand that's better for the rest of us. Less competition, lower difficulty, more chance of winning lottery payouts.

It looks like someone dug up a whole bunch of old clams from the initial distribution, and let them all stake.

Old clams stake relatively easily, because chance of staking is proportional to age times value.

One they all staked once, their age will have been reset to zero, and so the rush of staking abated (until the next person digs up a bunch).

Weight is just the sum of (age-in-days times value-in-clams) of all outputs currently trying to stake.

Each old distribution output is 4.6 CLAMs, and about 74 days old. That gives each one a weight of 340 CLAM-days until it successfully stakes.

If the network weight went up by 200k, that suggests that around 600 (200k / 340) old 4.6 CLAM outputs were found in a single dig!

Edit:

That matches pretty closely with the number of addresses that have been claimed. A couple of days ago the IRC bot reported:


And now it reports:

The formula is in the javascript source of the site, visible to everyone at https://primedice.com/js/init.js


Basically:

multiplier = 99 / chance