Page 1 of 1
Status as of Thursday, April 2
Posted: Thu Apr 02, 2015 9:19 am
by Steve Sokolowski
- Last night's database preparation for the upcoming release finished successfully. There is so much data that, even after deleting old shares, it still took 7768 seconds to add three columns to shares. Because of this prep work, actually releasing the new features will be much quicker.
- Right now, the only thing we need to do before we can certify this next release as working is to get access to kires's miners (the info he provided didn't work). Once we can make sure that his miners have the same behavior as our testing software, we should be good to go.
- Chris is working on one-day payouts, which would resolve some of the customer questions he often receives about the current 2-day payout system - but this won't be available until after the mining server release.
- We may release on the evening of Sunday, April 5, but that depends on how soon we can test kires's miners and whether any additional bugs come up.
- We apologize for the daily 2-minute restarts because of the memory leak in our existing code that will continue until the next release.
- After the release, it's possible that there could be emergent behavior that we were unable to replicate (because we couldn't actually find blocks at that 600GH/s fake mining rate). Because we can't afford to buy every model of miner, we have absolutely no idea what the average work restart delay is going to be. As far as I can tell, accounting for work restarts has never been attempted before, so it's going to be interesting to see what happens here.
- A lot of discussion has been happening about SHA-256 mining between Chris and me. SHA-256 mining is extremely risky, and could lead to tens of thousands of dollars in losses due to bad luck. We do want to do it in the future, but X11 mining might be a good intermediate step.
Re: Status as of Thursday, April 2
Posted: Thu Apr 02, 2015 10:40 pm
by loszhor
In terms of SHA256 being risky, is it a profit to usage situation or that there is not enough profitable coins to mine for the service space? If ProHashing does not implement SHA256 is it is fine with me as I have other alternatives for me miners but I'm just curious about the details.
Re: Status as of Thursday, April 2
Posted: Fri Apr 03, 2015 12:15 am
by Chris Sokolowski
The risk of implementing SHA-256 is that we are a pay-per-share pool; we pay users regardless of whether or not we find blocks. We need to keep a reserve of about 10x what the most valuable block is worth so that a string of bad luck does not bankrupt us.
For scrypt, the most valuable blocks are Litecoin, about $250 per block when they were the most expensive, and we we can easily keep $2500 on hand for bad luck. However, SHA-256 mining would require us to hold about 10x the value of a Bitcoin block, about $6250, for a reserve of $62500. While we could even afford that, the bigger issue is what happens when Bitcoins go on another big bubble. When they go on a run, then we need to acquire money quickly to keep our reserve steady. If they approach $5000 as Steve thinks in the next bubble, we would need a reserve of $1.25 million, well beyond what we own and what we can afford right now.
On the other hand, if we can grow the scrypt and X11 side of the pool first, then we can earn money to build up a reserve so that we can later fund SHA-256 mining.
Re: Status as of Thursday, April 2
Posted: Fri Apr 03, 2015 11:27 am
by Steve Sokolowski
Chris, I'm not sure that the issue of a bubble is a primary one, because that can be eliminated by simply buying 250 bitcoins. Then, it is irrelevant what the price of bitcoins is, because we will still have a reserve, even if those bitcoins become worth $1.25m. However, if bitcoins do rise that high and we only owe miners 50 bitcoins while owning 250, I think most people would agree we should just shut down, pay miners their previous debts, and live the good life rather than risking that reserve to earn more money.
On a different note, though, I think that X11 is probably more profitable than bitcoin mining. We can charge 5% or more because we have better technology, especially after the next release, that earns miners more than the losses in fees. In SHA-256, all of our switching algorithms would be useless, and we would just become another bitcoin pool in a market flooded with competitors. While we do have Coinbase payouts and the ability to be paid out in altcoins, I'm not sure if that is enough advantage over giants like GHash.io.