Page 2 of 2

Re: AuxPoW Proposal: Soliciting community feedback

Posted: Fri May 25, 2018 11:12 am
by GregoryGHarding
any movement on this?

Re: AuxPoW Proposal: Soliciting community feedback

Posted: Sun May 27, 2018 1:11 pm
by moonshot
GregoryGHarding wrote:any movement on this?
The NewYorkCoin dev team is pushing ahead.
There is no public branch at this time.
I'll post an update once the dev team has deployed on the testnet.

Re: AuxPoW Proposal: Soliciting community feedback

Posted: Wed Jun 06, 2018 6:54 pm
by moonshot
Steve Sokolowski wrote: The pool targets the assignment of dynamic miners so that no more than 50% of the network is mined by us. In addition to avoiding control of any one coin, it also doesn't make sense for a single pool to mine more than 50% of the blocks. Otherwise, it just pushes up the difficulty of its own blocks, causing the coin to become less profitable. It mathematically makes more sense to mine less than 50% of the network and assign extra miners to slightly less profitable coins.

The remaining 7% of miners were solo miners, which we can't limit because they can just decide to switch to another pool when they want to mine NewYorkCoins.

Other pools, especially single coin pools, probably don't have the 50% limit in their code, so they simply mine as many blocks as possible and push up the difficulty as a result. That's why you see more volatility in those periods.
Steve, the algorithm doesn't appear to work as described:

Image

Live View:

https://stats.nycoin.info/d/000000002/n ... 0s&orgId=1

Re: AuxPoW Proposal: Soliciting community feedback

Posted: Wed Jun 06, 2018 8:23 pm
by CSZiggy
why bother, no one cares.

Re: AuxPoW Proposal: Soliciting community feedback

Posted: Thu Jun 07, 2018 10:06 am
by Steve Sokolowski
moonshot wrote:
Steve Sokolowski wrote: The pool targets the assignment of dynamic miners so that no more than 50% of the network is mined by us. In addition to avoiding control of any one coin, it also doesn't make sense for a single pool to mine more than 50% of the blocks. Otherwise, it just pushes up the difficulty of its own blocks, causing the coin to become less profitable. It mathematically makes more sense to mine less than 50% of the network and assign extra miners to slightly less profitable coins.

The remaining 7% of miners were solo miners, which we can't limit because they can just decide to switch to another pool when they want to mine NewYorkCoins.

Other pools, especially single coin pools, probably don't have the 50% limit in their code, so they simply mine as many blocks as possible and push up the difficulty as a result. That's why you see more volatility in those periods.
Steve, the algorithm doesn't appear to work as described:

Image

Live View:

https://stats.nycoin.info/d/000000002/n ... 0s&orgId=1
This is because there are static coin miners who are mining NewYorkCoins here. These are miners who, if we didn't offer the service, would just set up their own wallets, so we can't really limit them. The dynamic coin miners will never exceed 50% of a network.

The increase recently is likely due to changes we made to the "live coin status" data. It's against our terms of service to use the data in that chart with bots, but we discovered that there were more subscriptions to the channel than there were website viewers. The data was changed to make it unusable for bots, so what appears to have happened is that people who were illegally using that data are mining single networks for longer periods of time. They have likely discovered that NewYorkCoins are profitable coins and have pointed a lot of hashpower towards them.

But I do have a concern with how this image is created. Someone asked us how to track blocks we mine the other day. We ourselves store all the block data for tax records, so if we ever get audited we can send the 10TB of data to the IRS and let them have at it. This isn't secret data, it's just too expensive to store it all on fast disks online, and we can't afford to run queries on it for people who ask without a large fee.

The data in the database is the only authoritative source of our blocks because we select random text to place in the coinbase transactions of blocks rather than a signature. The purpose of that is so that when we resend an existing block to miners, they don't think it's the same job and send duplicate work. A side effect is that there isn't a signature that makes it easy to track our blocks, and therefore the best this chart could do is to make an estimate, I think.

Re: AuxPoW Proposal: Soliciting community feedback

Posted: Fri Jun 08, 2018 4:41 pm
by GregoryGHarding
i have 2 gh on NYC for a few weeks

Re: AuxPoW Proposal: Soliciting community feedback

Posted: Wed Jul 25, 2018 11:36 am
by moonshot
Hello all,

After a few months of hard work we are happy to announce that work on the 1.3 wallet version is now ready for public consumption!

Key updates for this version:

--Supports upcoming security fork at block 4.8 million (Late October) - This includes a new difficulty adjust algorithm (Digishield) and enables Auxiliary Proof of Work (AuxPow/Merge Mining) support for mining pools.
--Now includes seeded DNS functionality - This will automatically find nodes based off of current network activity
--Several key security updates to help prevent DDoS and 51% attacks
--Extended key headers (brings compatibility to Electrum implementation)
--Block reward reduction starting at height 4.8 million
--New features for network and peer view in the Debug windows

As always PLEASE PLEASE PLEASE make a backup of your wallet.dat file (located in your data directory) BEFORE updating the wallet. We have had great success in updating our own instances but take no chances!

This version will use the same data directory as your previous installation unless you are using a custom data directory (in which case you should use the command line -datadir=<yourDataDirectoryPath> option).

This update also includes updated dependencies on many things such as the database so it will have to re-index the block data. I personally let it sync from 0 and it took about 14 hours. Reindexing should only take around 4-5 hours (this is HIGHLY dependent on your computer setup, things like SSD will improve that).

Things to keep in mind:
--Advanced Checkpointing will be implemented in a future version of the wallet, likely in the coming months
--The User Interface is not 100% where we want it yet and is subject to change.

Latest bootstrap and more wallet instructions can be found here: https://blog.nycoin.community/wallet-help/

https://github.com/NewYorkCoin-NYC/nyco ... s/tag/v1.3