Thanks.  I was thinking after I sent the bug report that it would have been smarter to have played some 1 BTC per game throws, waited until you paid out a few 2-2 and didn't pay out some 2-1 rolls, and then complained that the 2-1 winning rolls didn't get paid out.  You would have made good on them I'm sure, and the odds would have been in my favour.  Then I thought maybe others would do the same after seeing my bug report.  I hope it didn't cause you that kind of trouble.


OK, so you don't need to see the screenshot I just made of being paid 8.75 for winning the "any craps" bet...  I played the hand before going out this morning and only just got back to screenshot it.  I expect you've fixed that one too now.


You're welcome.  It's refreshing to have my bug reports appreciated, acted upon, and rewarded.  Rather than having the site owners ban me (sapphire casino), or ask me to stop posting (btcdice).  A site that gets mad when you point out their errors rather than seeing it as an opportunity to improve isn't a site I want to be playing at.

I don't have time right now to lose the 1 BTC back to you, but hopefully soon I will...  

Edit: your post doesn't seem to explain why 2-2 was being treated as a winning roll when betting the 1-2 bet.  I guess that was just a typo, right?

I played the two "17 to 1" bets (1,2 and 5,6) some more.

5-6 and 6-5 both pay out 17.5 chips.
5-5 is a loss.

I got 2-1 again, and it was counted as a loss again.

1-1 pays out 35 chips.  It's labelled as "34 to 1", so that's right.

So the "17 to 1" bets should pay out 18 chips, but they only pay out 17.5.

So the two bugs seem to be:

 * the "1,2" 17 to 1 bet pays out on 1,2 and 2,2 rather than on 1,2 and 2,1

 * the 17 to 1 bets both pay out 33 to 2 not 17 to 1.

1.  I've read about how to play craps several times before, and played it on various sites, but each time I come back to it I've forgotten what all the bets mean.  Some basic introduction to the game would be helpful.  But I guess that's not what you do - none of your other games have 'help' pages either.

2.  My first impression was that the font used on the table clashed with the usual bitZino sans-serif font.  It looks like it may be the browser's default font, and that the CSS styling isn't working.  Or maybe that's the font you chose on purpose.

3.  I was betting the 4 "single rolls" bets on the left each spin.  I rolled 2-2 and was paid out 17.5:


Then I rolled 1-2 and again was paid out 17.5:


Finally I rolled 2-1 and lost:


Is that a bug?  It's paying 2-2 but not 2-1.  And it's paying 17.5 for a 1 unit bet when the bet is marked as "17 to 1".

I imagine this depends on how many of his creditors default on their debts to him.

I just today received 20% of my principal back from Patrick, and over the last 4 weeks have received a total of almost 40% back.  I don't see why he would continue paying back his lenders if he was running a scam.  It looks to me like Patrick is trying his best to repay what he owes when he can.

That's very Sensible.

I'll post that information when I get a chance.  Sorry for the lack of a stats update yesterday.  Here's today's:

  I had pre-generated a bunch of vanity addresses and picked that one for the WBX funds.  Then after reading the recent "I've just been robbed" thread, it occurred to me that I'm not 100% certain that I've never stored the unencrypted private keys for my vanity addresses on my hard drive.  So I panicked and moved everything out of my vanity addresses.  Better safe than sorry!


Got it, thanks.

Imagine you send a bet and get nothing back.  Did you lose, or did your bet get lost or stuck somewhere?  You can't tell.

So they send a small amount back when you lose to let you know that the bet was processed successfully, and that you lost.

That way you're not left wondering whether their server is just being slow.

I just read what http://www.satoshidice.com/bits.php has to say about it:


which looks like a typo to me - perhaps it should say 'they return your bet'?  Either way, the losing payment is half of 1 percent of your bet.

The answer to your question will depend on what bitcoin client you're using.  The satoshi client doesn't gather up the smallest outputs to make a payment.  It will use a single output if it is just the right size.  Otherwise it will try to make the exact amount using the available outputs.  If it can't do that, it will try to make the smallest amount that is at least 0.01 btc bigger than the amount you asked for.  And if it can't do that, it will use the smallest output that is bigger than 0.01 btc more than the amount you asked for.  It's kind of a complicated algorithm but it works quite well.

If you use 2 separate wallets (one for 'big' outputs and one for 'small') your small wallet will end up with a bunch of tiny outputs that you'll almost never be able to spend without incurring a transaction fee.  If you mix the two together then the client will sometimes combine a big output with a bunch of tiny outputs.  The big output will 'carry' the smaller outputs and allow them to be spent without a transaction fee.

The way the fee works is this:  for each output you're trying to spend, multiply the value of the output in BTC by the age of the output in days.  Add up all those results, and divide by the number of outputs you're spending.  If the final result is greater than 1, then no fee is required.

So if you have a 1 BTC output that is 4 days old, and a 0.00000001 output that is 2 weeks old, you get this result:

( (1 * 4) + (0.00000001 * 14) ) / 2 = (4 + 0) / 2 = 2

So the transaction needs no fee.  The 4 day old 1 BTC output is old enough to 'carry' up to 3 worthless dust outputs.  But if the 1 BTC output was in your 'large' wallet it would never get mixed in with the dust outputs, so they'd never get spent for free.

A friend accidentally formatted my camera's memory card a few days ago.  I was able to recover all the files from the formatted memory card using "photorec" from Ubuntu's "testdisk" package.

7

After playing some satoshidice I had a lot of 'dust' in my wallet.  I made a transaction that combined a bunch of 0.00000001 outputs with some old large outputs in an attempt to tidy up my wallet a little.  The satoshi client determined that no fee was required, and so sent the transaction without a fee.

The miners seem to disagree.  The transaction isn't getting into blocks.  I expect if I wait long enough it will happen, but I'm wondering whether there's any way to cancel the transaction and send it again with a fee this time.

I tried using the raw transaction features of the new client, but when I try to send the raw transaction I get an error:

  error: {"code":-22,"message":"TX rejected"}

Presumably it's detecting the double-spend.

I tried using a new empty wallet which allowed me to send the raw transaction, but it's not showing up on blockchain.info so I guess now the network is rejecting it rather than my client rejecting it.

Great Freudian slip there Goat! 

Hi.

I've not made any further steps towards getting the repayments to happen.  I didn't have time to contact ITSA yet.  That's something I think I need to do first.  I've been very busy recently, but will get to it soon.  If they advise that I should pay back the WBX creditors then I should be able to do that soon.

In other news, I'm about to move the funds to a different address.  The address they're currently at was generated using vanitygen.  I figure it's possibly safer to use an address generated by the bitcoin client itself.

I'll be moving the 1768.9876932 BTC from 1DodGYjF5LFoKg186jANZV9U8znXjagSzS to 1BGjvqSQWgXtjZk1A8nda9b5HqBbecNHoF in the next few minutes.

Edit: I forgot to answer your 2nd question.  A total of 9 creditors have contacted me since I asked for contact details.

If we look at the historical data, June would have been a zero payout month, and July only just managed to recover June's losses too.  I don't know if these were freak occurrences, but so far 1 (and almost 2) out of 4 months would have been zero payout months:

OK thanks.  It sounds like the highest ever end-of-month profit is used to calculate this month's new profit, which makes sense.

As an example, if the price of bitcoin was the number being used instead of SD profits then there would have been no new dividends paid until the price goes above last summer's end-of-month high of around $16 ( http://bitcoincharts.com/charts/mtgoxUSD#igDailyzczsg2011-05-25zeg2011-07-05ztgSzm1g10zm2g25zv )

When you plot a log curve on a log chart, what does that look like on a linear chart?  Does it mean linear growth?

All exponential growth shows up as a straight line on a log chart doesn't it?