Quote from: yellowduck2 on June 15, 2014, 11:40:37 AM

Maybe u should open Nvidia multipool switcher using what u have created and charge pool fee. I bet u will make a lot of BTC in long run since Nvidia might take over AMD by year end and your pool will be unique with no competitor. Perhaps part of the pool fee goes to cbuchner1 as donation for being the creator of ccminer and continuous effort in improving ccminer. Developer win , you win , Nvidia miner win , everybody win Cool

Quote from: yellowduck2 on June 15, 2014, 11:53:24 AM

Quote from: cbuchner1 on June 15, 2014, 11:47:44 AM

I'd welcome the idea of someone creating a profit switching nVidia-centric multipool.

Patches to support such feature within ccMiner would be welcome.

yeah. It's really a win-win-win situation. Once the pool is stable , it's pretty much a continuous passive income for the developers and pool owner. That is if the pool is willing to give part of the fee as donation to you. Improving ccminer will improve miner profit and therefore increase income from pool and thus increase your passive income.

I built pretty much exactly this for my own use, but done slightly different.

It does all the calculations of the profit per coin per algo on the backend instead of on the client end. There is a simple client that is run on the client side which checks with the server every X minutes. If the server hands back a different pool/port then it closes the current running program and starts up a new program. I'm currently using this for both AMD and nVidia GPUs as well as for CPU mining. No changes needed to any mining programs.

I'm doing much, much better profit wise then I ever did managing it on my own or using any other multipools. HOWEVER, I'm not running my own pools since I don't have enough hash to make that worth while.

I COULD probably do this if there was enough interest but honestly I'm not a LINUX guy and would have to struggle my way through it since all pools are run in Linux and none under Windows that I know of. I'm a windows programmer and architect so I typically like to stay working in the environment that I'm comfortable in.

If I were to do this I'd probably just take it the extra step and do autoconversion to btc and payout in that manner.

The downside and reason I didn't really do any pools myself is that we'd then be competing with yet other pools. For example jpc.hashatme.com currently has 11.8GH of the network 17.71GH. So when mining JPC I'd rather be using a pool with 12GH then 500MH for example. This way when the "auto changer" does it's thing and for example switches to TAC the shares on the JPC are still good and will be paid out shortly (since total mining of the coin did not stop). If it were our own pool then all mining would stop and no reward paid for JPC until it gets mined again at a later time as there would be no blocks/shares found if no one is mining it. Typical multipool behavior.

If anyone knows of a proxy stratum that can be used in the middle to point to other pools I'd consider going this way and modifing it to track shares.

Quote from: djm34 on June 15, 2014, 02:17:51 PM

10% for a pool Shocked

Sorry to say but your idea is toxic Grin

(are you an xpm aficionado ?)

Also once everybody has the advantage, there is no advantage for anybody, so why pay 10% for something which won't be profitable anymore ? (this is where someone says: if you don't like it you don't mine it... but still... Grin

).
Just look at xpm, too many xpm have been mined now it is worths nothing (I know I forgot to sell mine Grin

)

10% for a pool is crazy. However I could see certain situations where a high fee would be appropriate for certain algos.

Take for example if Chris was to release a CryptoNight miner that could only be used via the pool. Here's an example using an AMD 280x single card rates:
SuperCoin can generate $1.37 per day for that card where Monero can generate $2.82. These are the two highest coins right now as I type this using different algos for the AMD 280x.
So if using that as an example. If the best we could do without a new release from Chris was $1.37 per day but now he releases to us a special new miner that can do $2.82 then that's double the profit. Would you be so stingy to not reward him with 10% of the profit you now get because of him?

I could see doing it this way. $2.82-$1.37=$1.45 additional profit due to new algo release. I could see the pool paying 50% directly to Chris on the PROFIT and 50% to the miner for this algo only.
So Chris receives $0.725 and miner receives $2.095. $2.09 is better then $1.37 which is all you'd get without his new release. So miner should be happy. Chris gets to make a little extra on each new release by using our hardware for a period of time.
That $73 cents per card would add up quickly for Chris. It could change the way he does releases. Instead of him having to manage hundreds of Amazon machines to make profit on his new algo before releasing it he could release it to the masses right away and make good money on all us nVidia miners. The 50% could drop 1% each day for example.

Probably a poor example above because I'm sure Christian's version of CrytoNight is much better done then the AMD version so profits would be higher for both parties.

Yes, global hash rate would go up for everyone. But in most cases Chris does release better algos then the AMD guys so we have a big advantage over the AMD/CPU guys in hash rate. Either way the above formula would only pay out the difference in profit so it's still fair for everyone.

It wouldn't matter if the hash rate is coming from hundreds of nvidia miners, hundreds of AWS instances or a combination. However, this could benefit Chris because the cost of AWS is higher then the pure hardware cost. There are plenty of times when the cost to mine some coins via AWS is not profitable while for us miners who already have the hardware it is still plenty profitable.