Worth discussing. Definitely a more long term goal
Primary Advantages
P2Pool
I think the biggest advantage would be allowing miners to run on a P2Pool node and even out variance quite a lot, which would be especially nice for our dedicated miners on LTC/VTC, and possibly DRK.
Running on P2Pool might also be valuable to the PS ports in certain situations. Although generally this would be counter intuitive, because of P2Pool's long PPLNS ramp up, for some coins (like LTC for example) it is often the most profitable but not really because of changes in difficulty or price. Which means mining on p2pool would even out some of the variance, without really reducing profitability.
PS between multipools & rentals
Pretty regularly speculators on rental services like NiceHash will bid the price of hashing power up far higher than anything that can actually be obtained via mining + exchanging. It makes sense to take advantage of those price jumps for PS ports.
Primary Challenges
Payouts/Accounting
Would require significant rethinking. Currently we use round based payout systems, but most proxied stratums will be paying out based on shares submitted. For things like P2Pool payouts might still be coming in for shares submitted several days ago.
Feedback/looping
What happens if we point to nicehash, and then they point hash to us?
Latency/Stales
Geolocation becomes more important in this sort of setup, otherwise latency can stack up pretty high
Worth discussing. Definitely a more long term goal
Primary Advantages
P2Pool
I think the biggest advantage would be allowing miners to run on a P2Pool node and even out variance quite a lot, which would be especially nice for our dedicated miners on LTC/VTC, and possibly DRK.
Running on P2Pool might also be valuable to the PS ports in certain situations. Although generally this would be counter intuitive, because of P2Pool's long PPLNS ramp up, for some coins (like LTC for example) it is often the most profitable but not really because of changes in difficulty or price. Which means mining on p2pool would even out some of the variance, without really reducing profitability.
PS between multipools & rentals
Pretty regularly speculators on rental services like NiceHash will bid the price of hashing power up far higher than anything that can actually be obtained via mining + exchanging. It makes sense to take advantage of those price jumps for PS ports.
Primary Challenges
Payouts/Accounting
Would require significant rethinking. Currently we use round based payout systems, but most proxied stratums will be paying out based on shares submitted. For things like P2Pool payouts might still be coming in for shares submitted several days ago.
Feedback/looping
What happens if we point to nicehash, and then they point hash to us?
Latency/Stales
Geolocation becomes more important in this sort of setup, otherwise latency can stack up pretty high