Topic: ethminer-0.9.41-genoil-1.1 - page 35. (Read 397363 times)

if you use the miner in parity maybe, this is a standalone miner that would connect to parity via RPC

I read on another miner thread that parity brings us  new more efficient miner so that 's why I am asked

Geth and Parity are the same thing, and Ethereum client with RPC support.

Genoil's miner version should support both with getWork (-F), so, I suppose it's not really up to Genoil.

The fix is don't use Windoze.

 

Dear Genoil,
Will you  build new version still on geth or on Parity ?

I don't even know about a V3.
The way I understand it, there's the dwarfpool one, then coinotron's based off the bitcoin/slush protocol, and now nicehash. I don't know about version numbers except nicehash adjusted the coinotron one to allow for their needs, so if coinotron's was V1, I guess nicehash is V2.
This was the info nicehash gave http://forum.ethereum.org/discussion/7369/dagger-hashimoto-added-to-nicehash. Other info I found was from https://github.com/Coinotron/eth-stratum-mining-proxy/commit/065bf6423a310a590029cf2f2395c7f1e21467af for the original coinotron stuff.  For the dwarfpool stuff, I read through the code https://github.com/sammy007/open-ethereum-pool but that's probably not the easiest to pick through. Wish someone put out a full spec.

there isn't, but try this if you can


i think on win 7 new pascal are doomed for etheruem, there will not be any good hashrate there

What is the fix for nvidia 1080 on windows 7 64bit? I am getting .75MH/s, lol.

Yup, other's who aren't paying attention.
ANY share can solve a block - it's stupid not to allow this and just look for duplicate nonces and verify the share diffs.
sammy007 is convinced this is a bad idea and his pool doesn't (and never will) support vardiff.
He also won't share why he's so paranoid on the subject.

I've examined the code in geth and I just don't see a problem with a difficulty "proxy".
It's not REALLY vardiff, just a mech to balance share submission.

Not true with DonkeyPool.
Nnginx sits between the wallet and proxies eth_submitWork.

The getWork params only change when there's a blockchange, as it should be.

Stratum is marginally better - for pools.
It's easier to administrate - but keeping a socket open has essentially the same footprint as 2 requests to nginx per second, per miner.
There is no spec, only code.
NiceHash's document is outdated and I haven't found it to be accurate.
Maybe they just haven't updated.

Try DonkeyPool - min diff is 1.5 Bill and fluctuates so all miners share submit average is equal to the average ETH blocktime.
That's what stratum is suppose to do, right?

There are now (3) stratum options (thanks to NiceHash).
They each have differing opinions on how to deliver getWork from the wallet without polling a websocket.

DonkeyPool supports the Coinotron spec, which makes the most sense.
Why use a proxy? all miners now support some type of stratum.

DonkeyPool still has getWork support, but only until June 30.
Then it's all stratum(ish) support.

DonkeyPool getwork string would be address: / workername / hashrate integer value

eg
www.donkeypool.com/0xaaaadddreeesss/worker/10

You should utilize port 3001, 1 Billion Difficulty.


Port 3000 is useful for lower hashrate miners, such as those mining on a single GPU.

Port 3001 is the recommended port currently if you are not sure, especially since a majority of purpose built mining rigs are around the 80-150 MH/s mark currently. Even if you have several workers, they are still separate miners.

Port 3002 can be utilized as well, it is set at 2 billion difficulty and is a standard at other pools.

Further ports are meant for users that are renting hashrate, where all the workers are funneled through one proxy and appear as one worker. Also users that are utilizing Sammy007's proxy etc.

Generally speaking, multiple static diff ports was previously a taboo thing for an Ethereum pool to do. Before we investigated and fixed the issue, malicious miners used to be able to submit the same block header hash across multiple ports to scam an unsuspecting pool. Our development team figured out a way to patch this to permanently prevent that from happening.


We also currently have an experimental version of Variable Difficulty (VARDIFF) we are working on. We had to re-invent the way VARDIFF works for Ethereum. It cannot be compared to a Bitcoin/Bitcoin Clone type of pool that utilizes VARDIFF. The current VARDIFF system will re-target your miner every 10 minutes, trying to find the perfect difficulty so that you are submitting a share every 10 seconds. It may take a few hours until that diff point is found, and currently if you disconnect or timeout (30 minutes) you will have to start over again. That is why we offer several different starting difficulty points for VARDIFF.

Variable Difficulty is still in experimental/testing stages. Use at your own risk.


VARDIFF Ports:
4000 - 300 Million Starting Diff
4001 - 1 Billion Starting Diff
4003 - 3 Billion Starting Diff
4005 - 5 Billion Starting DIff

lower hashrate = lower diff

I am curious which diff is recomended, too..

Can anybody with a 1080 or 1070 and Win10 try to get 368.20 drivers running? They are installed by Windows Update and may require a Windows Insider subscription. When running dxdiag with 368.20, WDDM version is upgraded from 2.0 to 2.1. No TLB fix for my 970, but it may solve the issue on Pascal cards.

I don´t ask for a pool, but for several diff ports of Alpereum.

i recomended dwarfpool im use it too.

noob question - what is the sense of these static diff ports?
Which one is recommended for ~88-90 MH/s?

Hello,  flags changed a Little  in  new version

It should be - SC instead of SV now