Survey of new features

We've had some feature recommendations and I thought it would be worthwhile to post a poll to see what sort of demand there is for them. All of these features require significant time, so working on one would delay the others.

Please vote and reply to let us know what you think!

Steve Sokolowski wrote:We've had some feature recommendations and I thought it would be worthwhile to post a poll to see what sort of demand there is for them. All of these features require significant time, so working on one would delay the others.

Please vote and reply to let us know what you think!

My vote is simple.. increase profits!

Another feature I would love to see would be to add myself coins in which I would like to be paid out! Not sure about feasibility but at least I gave it a try

After the AI for deep learning I'd like to see X11 through X15, Qubit, and Quark support added as I'm considering buying some miners for those Algos.

gaanthony wrote:After the AI for deep learning I'd like to see X11 through X15, Qubit, and Quark support added as I'm considering buying some miners for those Algos.

If you mean a Baikal miner, I would suggested you didn't buy. I have one, works great. Not a single problem with it in weeks, BUT the profit has dropped from ~$7usd a day down to ~$2usd a day almost overnight. The $500usd purchase price is simply too high to make it worthwhile. That is a 250+ day to ROI and it is getting worse each day as they sell more miners. BTW, very nice people at Baikal too. My only issue is PROFIT, and they simply are not profitable at the $500 price point.

Some1notu wrote:

gaanthony wrote:After the AI for deep learning I'd like to see X11 through X15, Qubit, and Quark support added as I'm considering buying some miners for those Algos.

If you mean a Baikal miner, I would suggested you didn't buy. I have one, works great. Not a single problem with it in weeks, BUT the profit has dropped from ~$7usd a day down to ~$2usd a day almost overnight. The $500usd purchase price is simply too high to make it worthwhile. That is a 250+ day to ROI and it is getting worse each day as they sell more miners. BTW, very nice people at Baikal too. My only issue is PROFIT, and they simply are not profitable at the $500 price point.

Thanks for the feedback. I've had the same experience with bitcoin miners - KNC Jupiter, Neptune, Bitmain, then scrypt with KNC Titan. Have also tried cloud mining contracts with no ROI. My thoughts are around long term algorithm diversification. I like that the cost to run the Baikal and related X11 miners is way way less than what I spend running KNC Titans currently. I understand ROI as long gone are the days of 90 day or less ROI. Any investment is a risk. Just like gambling don't invest any more than you are willing to lose.

galixxx wrote:Another feature I would love to see would be to add myself coins in which I would like to be paid out! Not sure about feasibility but at least I gave it a try

Do you mean that you want to be able to enter the name of a coin and have it automatically pay out in that coin?

Steve Sokolowski wrote:

galixxx wrote:Another feature I would love to see would be to add myself coins in which I would like to be paid out! Not sure about feasibility but at least I gave it a try

Do you mean that you want to be able to enter the name of a coin and have it automatically pay out in that coin?

Yes, I mean that exactly. Eg. I love the coin OKcash. Having a direct payout from prohashing.com would help me. Currently I am either paying in BTC or other coin (I would exchange the coin in BTC) and then would be able to get OKcash.
Not sure if I am the only person who had this thought?

galixxx wrote:

Steve Sokolowski wrote:

galixxx wrote:Another feature I would love to see would be to add myself coins in which I would like to be paid out! Not sure about feasibility but at least I gave it a try

Do you mean that you want to be able to enter the name of a coin and have it automatically pay out in that coin?

Yes, I mean that exactly. Eg. I love the coin OKcash. Having a direct payout from prohashing.com would help me. Currently I am either paying in BTC or other coin (I would exchange the coin in BTC) and then would be able to get OKcash.
Not sure if I am the only person who had this thought?

Thanks for the clarification.

Unfortunately, I don't think we'll ever be able to provide the automatic addition of coins, because adding a coin is a time-consuming process that can only be done manually. Chris needs to search the Internet for the right source code, download its blocks, and add the appropriate rows to the database, which takes many hours. Since every coin is different, we can't just write a program to add a coin. It is also costly to add new coins.

However, you can ask Chris to add a coin by sending an E-Mail to contact@prohashing.com. As long as the coin sustains $2 in payouts, he will maintain it for you.

Please vote on this poll! The votes will close soon.

So far, I'm really surprised at what has been found. The overwhelming majority of feature requests are for multiple algorithm support, but here X11 support has not finished first. We will probably need to go with multiple algorithms first anyway, simply because the codebase is already forked and work has started, but this poll will probably move any website improvements way down the list, as nobody has expressed any desire to see more charts.

I'll have more knowledge about what it would take to use machine learning within a few weeks. My guess is that rules-based coin switching is ideal for most cases because we can mathematically prove that it is right relatively easily. The AI, however, can be used to determine several things that humans can't understand:

• How can we switch coins to reduce our orphan rates?
• How can we predict the behavior of other pools and the effects those pools have on the networks?
• Can work restart times be better determined?
• Can the difficulty algorithm of a coin be learned by AI so that we can build upon the previous block without needing to call getblocktemplate first, because it takes far too much human intervention to enter this data ourselves to be profitable?
• Can it be determined when the bottom is about to fall out of a market?