2,147,483,647
There seems to be an error in the mpos code. This is a classic 32 bit code fuckup. That's the maximum a 32 bit signed integer can be. We long ago passed needing numbers much bigger than that in bitcoin... Yes, that's 2^31-1, a pretty recognizable number to most C/C++ programmers. Is it the pool front end capping it at that or are the pool share calculations doing it? It might pay to check if calculations are being done with 32 bit values anywhere. I would have thought it'd overflow rather than be capped at max though. I'm guessing it's just on the front end as the shares are all properly stored and accounted for... I'll be looking at the front end pages as well as the payout scripts to ensure accuracy. |
|
|
|
All, I have completed the spreadsheet and have uploaded it to the following location: https://www.dropbox.com/s/99ijt4by0jc8nq4/pool_payouts.xlsx?dl=0If you do not have Excel, or google docs, or any other spreadsheet program, please let me know and I will gladly present the data to you here on the thread. Looking at the spreadsheet, pay attention to column R. If that number is NEGATIVE, you were UNDERPAID. If the number is POSITIVE you were OVERPAID. Some of you have already committed to paying back any coins - and in fact, some of you have already sent those coins to me. If you are underpaid, and you wish to have things corrected, please let me know. I will send you coins from my own pocket. Of note: I was the worst affected of any miner on the pool. Even though I was overpaid for block 3, I was significantly underpaid for block 2. Net loss for me was 4.96 BTC. If you have any questions or comments, I will address them to the best of my ability. My GoJonnyGo account has -0.3329574 This is a little more than I'm willing to just let go so would be grateful if it could be sent to my payout address. Thank you. I will send as soon as I can... Been stuck in airports all freaking day. At least I was able to compile the spreadsheet while waiting for yet another delayed or canceled flight... |
|
|
|
All, I have completed the spreadsheet and have uploaded it to the following location: https://www.dropbox.com/s/99ijt4by0jc8nq4/pool_payouts.xlsx?dl=0If you do not have Excel, or google docs, or any other spreadsheet program, please let me know and I will gladly present the data to you here on the thread. Looking at the spreadsheet, pay attention to column R. If that number is NEGATIVE, you were UNDERPAID. If the number is POSITIVE you were OVERPAID. Some of you have already committed to paying back any coins - and in fact, some of you have already sent those coins to me. If you are underpaid, and you wish to have things corrected, please let me know. I will send you coins from my own pocket. Of note: I was the worst affected of any miner on the pool. Even though I was overpaid for block 3, I was significantly underpaid for block 2. Net loss for me was 4.96 BTC. If you have any questions or comments, I will address them to the best of my ability. |
|
|
|
Jonny, if possible, can you please check my payout too. Thanks.
Pretty sure he's going to go through everyone's to be safe I am doing everyone's. However, since he asked, here's the breakdown for eghigo: Block 1 - 0 shares Block 2 - 5575801325.551303 shares Block 3 - 15199188758.73043 shares Actual payout (before fees): Block 1 - 0 Block 2 - 1.95214418843209336352 Block 3 - 0.8998633233949909771 Expected 5N payout (before fees): Block 1 - 0 Block 2 - 1.23713195430087 Block 3 - 1.44418695028503 Sum of actual payouts: 2.85200751182708 Sum of expected payouts: 2.6813189045859 Overpaid by: 0.17068860724118 You actually made over and above what you should have. |
|
|
|
I'm actually working on a spreadsheet that will contain all of the numbers and will publish it when completed. I've just started and have gone through my own account... sure, I may have been overpaid this block (by about 0.6 BTC) however, I was underpaid on the second block by about 5.54 BTC. Somebody made out pretty well  . The answer will be made available when I'm done. Might be a day or so - I've got to travel to the client site tomorrow and the little lady's pretty sick of me spending Valentine's Day buried in the office doing that damn Bitcoin thing again. |
|
|
|
I think he can leave it the way it is.
then square it up with the underpaid.
You are an underpaid and rozo is an underpaid.
the biggest overpaid on this block is jonnybravo0311
I would be willing to do a manual check of all numbers this evening.
Actually, that's not entirely accurate. Kilo was definitely underpaid - I showed the numbers. Rozo was actually overpaid as were you and I. I'll gladly post the numbers if you wish. |
|
|
|
|
To keep things open...
kilo17 - total shares submitted to the pool is: 3593906827.6008587. Total shares for the 3 blocks found: 265158127564. Last block found was: 25.19462806BTC. Your total payout should have been: 0.34148358BTC.
If that is acceptable, then give me a BTC address and I'll send it to you.
|
|
|
|
|
Thanks for the reply kilo, I appreciate your response. As the pool operator, it is my duty to ensure people are paid according to how the pool was advertised. If I'm not doing so, then it is also my responsibility to make things right.
That's why I'm offering to make the corrections from my own pocket and will happily send you the .338BTC you should have received.
|
|
|
|
|
By the way... additionally, if you truly feel that you've been somehow wronged, let me know. I will pay you out of my own pocket.
|
|
|
|
|
The pool is currently calculating the payouts based on an average number of shares. That average is calculated by looking at the shares from the last 5 blocks. I incorrectly stated the PPLNS was 5N where N was the block difficulty. That is not how it's working now. This is the reason why you're seeing the weird share caps and the odd things on the stats pages. As of now, since the pool hasn't found 5 blocks, it is basing the PPLNS average shares on the previous 3 blocks. This is why you are seeing round shares, but not seeing any payouts. Your round shares were dropped from the PPLNS calculation.
There are a couple of options here.
1) We continue on as is. This round's payouts will be calculated the same as they have been.
2) I try to re-run everything setting the share target as 5N (which would include every share since the pool began). This would be a manual effort on my part and will take a few more hours to complete.
There are 3 more blocks to confirm before the pool sends the balances to miners. I can try to stop this from happening... but I don't want to cause any more problems than already exist.
Kilo17, your assessment that 5N might be too much is an accurate one for a pool of this size. The difficulty is changing faster than the pool is solving blocks - which works out well for the current method of PPLNS that is implemented. If the pool grew to where we were solving lots of blocks per difficulty change, then 5N would make more sense.
Having written this, I am open to hear thoughts on how people would like to proceed.
My feeling? Leave it as is for this block's payouts. I really don't want to start messing around with the database and potentially screw things up worse than they already are for people expecting 5N payouts.
If we all agree, then I can change it to really consider 5N for the next block - I've got all the shares stored in the database anyway.
|
|
|
|
|
Fair points, kilo... I'm looking into it.
|
|
|
|
Actually, I don't qualify for my own promotion... So that means the bonus is still available to the next block finder  . I will keep the promo running. Same rules as before: 1 BTC to the block finder until the pool crosses 110% of expected shares, at which point it goes to 0.9 BTC until it crosses 120%, then it becomes 0.8 BTC and so on. |
|
|
|
Wow... a ton of activity on the thread... you'd think the pool found a block or something! To those asking about payouts, I truly hope Phil has shown how changes in hash rate of the pool will have an effect on your estimated and actual earnings. Now I've got to think up a new promotion... |
|
|
|
|
Well my order is there but not active because of an infinite hash order at 38 and a 7PH at 39.
|
|
|
|
|
Phil, I'll join you. Just waiting on confirmations in my NH wallet and I'll throw 500TH up for about 12 hours.
|
|
|
|
I am glad someone agrees, Does the pool in your sig Apply this idea ?
No, my pool does not limit your hash rate in any fashion. I don't care if you've got 1 GH/s or 1 PH/s. The point of my post is that your hash rate expects to earn the same thing no matter if you are pointing it at a small pool like mine, or a larger pool. You can't game the system. I don't understand what you're hoping to achieve by offering a pool with a hash rate limiter on miners. Let's say you limit a miner's hash rate to 1TH/s. Fine. Now you get 10,000 miners with 1 TH/s. How is the end result in any way different than you being on a pool with miners of all different speeds, but the pool's total is still that same 10 PH/s? |
|
|
|
OK... it's been a while, and I think it's time for another promotion. The pool is about to cross the 100% of expected shares for a block threshold. There will be a block finder bonus for the next person to find a block on the pool. The reward decays as the percentage over expectations increases. It starts at 1 BTC and goes down as follows: 100% to 109.99% = 1 BTC bonus to block finder 110% to 119.99% = 0.9 BTC bonus to block finder 120% to 129.99% = 0.8 BTC bonus to block finder ... Obviously, the sooner you find a block, the better your reward will be . Oh, and as usual, I do not qualify for my own promotions. The promotion ends either when the block is found, or if the expected shares crosses the 200% mark. Good luck! |
|
|
|
I wrote this on another thread, but it completely applies here: It's statistics. You are correct that you'd see more per block here than on other pools. However, you would see far fewer blocks here than on the other pools. A pool with 20 PH/s expects to find 2.7916 blocks a day at current difficulty. A pool with 1 PH/s expects to find a block every 7.1633 days. (these numbers are rounded off)
Let's say you have 100 TH/s of gear. On pool 1, you represent 0.5% of the pool's total hashing power. All things being equal, you would expect to make 0.125BTC per block. On pool 2, you represent 10% of the pool's total hashing power. You would expect to make 2.5BTC per block.
Pool 1 would solve 20 blocks for every 1 block solved by pool 2.
So, what does that mean? Yup... you make more per block on the smaller pool; however, by the time you've found that block you've made the same amount of coin on the larger pool.
Now, when you factor in the nature of mining (luck swings, difficulty changes, orphans, etc) you are far more likely to be impacted on the smaller pool than on the larger one.
|
|
|
|
|
Isn't an "almost block" nothing more than the miner's share difficulty? The only way a pool has any idea of your hash rate and how much of the block reward you should receive is by tracking how many shares you submit. Almost block... lol.
|
|
|
|
|
Show Posts
