Hi Patrike,
Small requests as related to the above discussion if possible.
1) Include (non-dynamic) properties to set block time for coins (CTM happens to return 0 for a good number of them, and I'm being maths incompetent, would prefer method 2 above for calculation in Excel regarding possibility of solo mining)
1-1) (Not Important), most wallet JSON can report time elapsed since last block was solved (I think in seconds, or relative linux UTC timestamp), again, might be useful if this metric can be included for solo-mining (also display via AM HTTP API response (GET
http://mypc:17790/api/coins/stats method) and coins tab)
2) Include (dynamic) properties to fetch network hash rate via custom sources (again, CTM is either displaying 0 or mostly incorrect data as sometimes Yiimp pool APIs it uses report hashrate incorrectly), again, this is probably going to be useful metric for solo-mining.
***Really want to have feature***
3) Ability to set pool profit factor via HTTP-API
Best Regards,
Thanks for the suggestions.
For 1 and 2 I suppose we would also let Awesome Miner use the NetHash formula if for example the coin Difficulty is zero. Have you had a chance to test if this formula gives accurate numbers?
CoinsPerDay = hashRate / (hashRate + nethash) * secondsPerDay / blocktime * reward
3) Today you can only set Profit Factor for a Coin over the API, but not for a pool. If it's needed for pools as well it should be easy to add.
Hi Patrike,
I haven't tested the "accuracy" of the formula cause luck is also a factor + I used autoexchange on ZergPool which made it a little hard to track earnings, but it was what I was using for the past 2 days for experiment. (see pic)
https://imgur.com/HBOVlmxso from the picture, you can see I'd need to add my own BlockTime where AM's API reported 0 due to coin providers, then some factors I consider important as below.
* The NetHash has to be accurate, cause else it'll affect the entire calculation as per the formula we discussed here
* BlockWorth for solo as some blockchains block reward just won't cover the cost even if you win 100% of blocks even the difficulty was low etc. This this is filtered to suit one's need (for me, 500 satoshis per minute)
* BlockWinChance, Simply used the hashRate / (hashRate + netHash) to work out percent chance I'm likely to hit a block
* BlockWin_IntervalSec, to keep consistent flow of earnings a big win every few days/weeks may not be as good as lots of little wins daily
* SatPerMinute, Just for earning reference in perfect fair conditions (not sure I used the terminologies correct at all, but you get the idea)
* Division <--- now, this is something I'm trying to find a way to prioritize pools with a balancing factor but like I have mentioned, my maths are bad so I don't even know if this is right (currently using SatPerMinute / BlockWin_IntervalSec).
Result for the past 2 weeks follows, AM estimate on earnings are averaging about 40 USD or more, I tried solo mining, conventional shared mining, switching interval of 5~10minutes. regardless, I'm only averaging about 200 satoshis per minute, which is far from the earnings estimate regardless of which formula I used, call it a streaks of bad lucks, but even shared mining didn't end up much better, the Yiimp pools (used Zerg, zPool mainly) tend to seem to have particularly bad lucks that usually sees block efforts 500% regardless solo/shared mining.
This is to report on the effect of the Dynamic Coin Properties Update on earnings I promised 2 weeks ago since it was made available.
Now it's only been 24 hours since I used the Excel custom profit factoring to prioritise pool, so I cannot say this is going to continue but so far I'm seeing vast improvements on earnings (doing solo mode on most chains where feasible for my farm's hashrate on ZergPool with AutoExchange), I'm averaging approx 400 satoshis per minute which is about double of what I get for the past 2 weeks via various methods as described above.
***
So if this works out, you might consider some similar sort of mathematically correct feature to implement some kind of profit factors based on the above metrics (since they are already all available in AM) similar to how you apply factors for Online Services vs Actual/Estimated earnings, I believe this will be then be a lot closer to SMART profit switching, more likely better than what NPLUS Miner's claimed PLUS logic ;p
***
3) Currently I worked around the need to apply pool profit factors by using the API to include the top 3 pools in the Excel List I showed here in my example, so ability to modify pool profit factor via API is not that much sought after anymore, since there are also benefits for a small selection of profitable pools to mine for, as some of my hardwares are low memory (1060 3GB) and they won't work on many of the new algos such as MTP, GRIN...etc.
Best Regards,
I will look into this in more detail after the 6.3 release. Most likely Block Time and NetHash will be two configurable properties for a coin with support for Dynamic Updates.
In addition, the coins-per-day and revenue can then be calculated using the NetHash+BlockTime formula instead of the formula based on Difficulty. Either by a setting per coin or simply use the NetHash formula if the Difficulty is zero.
good decision, those two values through http api as we are using difficulty or price, will make the data much more realistic.
I am not a user as advanced as moppidoo, I will not use an external Exel and handle my own data. But if you read the apis and Json, and the power to add blocktime and nethash, you could give us something of value by choosing between the two possible ways of doing the profit calculation that you explained to me.
We in our group use an online excel and we have pointed the urls of difficulty, price, request of the price, profit of the currency etc ... We have noticed that when a currency its blocktime is higher than 1, the profit is usually lower . If the blocktime is 1, the value is usually quite close to what AM indicates. There are coins that we have to use profit of 0.65 so that the estimate of AM and what is obtained is the same.
To be able to add these two data and be able to use them for self-profit in some way, instead of its classic formula, could be interesting. Have two types of profit to choose, two different formulas. One based more on the nethash and blocktime, and the classic one of you.
One last suggestion Is there any way that all the data in AM can be saved from the last time data has been requested from the APIs? I explain. I use CC and make a request every 7 minutes and read all the data, but then I make requests from loose currencies defined by us that if we do not find the Explorer or we can not read the explorer, we use WTM, CC or the one we see more realistic. But then we exhaust the requests per day, because each currency that I ask for a data, is a request to the api.
If in each complete request of api of WTM, CC, CTM, Coinwarz, the last one is stored in AM, and that data could be read from the cache of the last time within the AM itself, the problem of excess requests to these services.
But that's basic, we also make requests to Exchanges to extract the price, if every 7 minutes I read 24 coins of the same exchange, make 24 requests, and in a few hours I am unable to ask for more data to that exchange. I do not know if once an API is read, whatever it is, it is saved in a kind of cahe, whether it is an exchange or a provider, and only save the last time it was requested, in my case every 7 minutes. So if I have to read several prices of different currencies of the same exchange, I read it in local because I have the copy of the received in the api.
I hope you understand my idea.
