#1450 Que lleve 4 años no quiere decir que no te venda humo con sus optimizaciones y que te este robando mas shares que ese 2% que dice.
A parte que usa los steal shares para mandar un falso hash rate y hacer placebo con los mhz que ve la gente en el miner, cuando en verdad no mina mas rápido que los demás.
Aqui esta muy bien explicado, incluso alguno ha echo reveres enginery.
Si todo esto es cierto la verdad es que es una puta basura.
Yeah that's pretty close, but he doesn't overreport current hashrate, or do anything more than find a way to submit a stale share for your worker for every valid share submitted during DevFee mining to his stratum connection. AKA for each share found in DevFee mining he gets 1.11MH/s current effective and you get 0.736MH/s current effective for 30 minutes.
(UPDATE: he does over-report hashrate, taking into account a timer used to determine the last time DevFee mined -- he over-reports while your not DevFee mining and when you start to get closer to when he want's to dev mine he drops the reported hashrate off so when you see your effective hashrate drop it's less obvious that it is caused by DevFee mining)
So he takes about 4% to 6% (because on-top of the number of seconds he mines you have to factor in handling that check/loop, opening the stratum connection, starting to mine on that PoW hash while still managing to report hashrate for your main stratum connection, then closing the connection (etc)) -- then by this method of giving you stale shares, he gives you back 2% to 3% ...
Either people connect/start their miners around the same time or there is some other reason, i know that one of the reasons they sync up is EVERY time the DAG updates he resets the DevFee timer so that every single active worker will mine at the EXACT same time period aka at 0:15 and at 0:38... (After first hour: all that extra load on the pool causes the lag, your hashrate drops way low because for 0:10 to 0:30 you don't submit any shares because your d/c then reconnecting blah blah (explained all this before, see links below)) ...
UPDATE: also confirmed that he updates another timer that seems to handle the very First DevFee mining timing directly after a DAG update or when first starting miner... this is the mechanism that lets most things recover by about 31:48 in that hour, he allows you to have some of your real shares and slightly boost your current effective hashrate right up until it start's DevFee and get's a connection open by about 38:50 ~ so you see hashrate going up on your stats page from 40:00 to roughly 50:00 before you realise it's dropped off massively by either now or when the next hour ticks over (or more specifically whenever the pool stops lagging from all the load). Net Result: You now all mine in the same time period twice every hour making your hashrates drop off and his spike up, and using stale shares on your stratum connection to make the drop off less obvious initially and causing these huge lag spikes and issues discussed in further links. The lag sort of makes people think they are probably stale shares and not something sinister.
This compounds it so he makes huge chunks of cash, because PPLNS will pay him for his current effective hash rate (which will hold for up to an hour at high high levels) -- meanwhile your hash rates are so low your earning half what you should, seemingly 2 times an hour every hour every day. (if there are other event's like an ico the extra load drags each of these cycles on for up to 2 hours)
GRAPHS: https://www.reddit.com/r/EtherMining/comments/6h01sc/is_it_worth_switching_from_genoil_ethminer_to/diuojr0/
MORE INFO:(For people wondering where the rest of the detail about what i'm talking about is):
https://www.reddit.com/r/EtherMining/comments/6h01sc/is_it_worth_switching_from_genoil_ethminer_to/diur2zv/
https://www.reddit.com/r/EtherMining/comments/6h01sc/is_it_worth_switching_from_genoil_ethminer_to/diugisx/
https://www.reddit.com/r/EtherMining/comments/6h01sc/is_it_worth_switching_from_genoil_ethminer_to/div0g10/
https://www.reddit.com/r/EtherMining/comments/6h25b8/is_the_claymore_miner_an_nsa_skunkworks_project/dix03m5/