Excellent work, but please explain the functionality of the scripts a bit more - a lack of any sort of explanation or anything but the most basic of instructions with Nemos' version has been very frustrating.
Also consider the following if not already implemented or in the works:
1. Increase the default interval for polling the API to 120 seconds. <- I use 300, for example.
2. Only obtain profitability info for the selected algos and the pool in use (unless there is a fallback pool specified).
3. Monitor miners so if there is a disconnect or benchmark failure the script moves on to the next algo, rather than just get stuck.
4. Provide a clean way of exiting. Right now I have to ctrl-C the powershell script, then exit the PS environment, then ctrl-C the miner then, even though there are no instructions to do so, I run stop.bat.
I have some other suggestions but I will wait until I've seen how your fork works before commenting further (a novel concept, eh?).
Thank you.
1. Will consider. Fixing Point 2 makes it OK I believe. One can change it in the bat file anyway.
2.
Fixed on master. Will be part of the next release. < This is what killed ahashpool plus servers. You have to learn the hard way sometimes:)
3. No plan yet. But i hear the ask.
4. I don't get that one. Closing NPlusMiner window should be fine.
MrPlus
Okay, I rescind my first suggestion as I am now all but convinced that a rolling average of the last 3-6 snapshots of profitability obtained at 60s intervals will achieve the desired smoothing of the data without losing the ability to respond to legitimate upswings in profitability; averaging the instantaneous and 24h data, which is what I have inferred is being done now, will not behave the same way as it will have the net effect of cutting volatility in half (assuming equal weighting of instantaneous and 24h profitability).
Suggestions 2 & 3 still pertain, but I think just requesting profitability data for the default selection of algos for only the pool in use is fine, rather than further restricting the request to only the actual algos specified in the start[pool].bat file. That may be less tedious for you to implement.
I also rescind suggestion 4 because simply closing the script window - rather than trying to ctrl-c out of it - works fine and doesn't leave any processes still running (which is what I was mainly worried about).
Now, an additional suggestion/caution - I see several, but not all, of the ccminer command lines have intensity specified. While it has been my experience that manually setting intensity can get a better hashrate than letting ccminer choose it automatically, it can also cause a worse hashrate or even crash the miner so I strongly suggest just letting each ccminer instance choose intensity for itself.
EDIT - one more suggestion, though a minor one, is to start each instance of the actual mining program minimized so focus doesn't shift to it. Palgin's neoscrypt miner is an especially egregious violator of this, rudely thrusting it's ugly face in mine whilst I'm trying to do something else. This would be less of an issue on a dedicated mining rig, but right now I am testing your script - and various other things - on my main desktop.