Quote from: cryptonewb on November 19, 2015, 03:05:15 PM

I am a bit out of the loop.

Is there any news on the Evolution front? Any details about how the mixing without masternodes will be implemented?
Any news on which other improvements Evolution will bring?

Still not bought a full masternode. It seems the price is a bit weak :/ (I lost 0.6 BTC on my 200 DASH)
I want to wait a bit until Evolution is implemented I think.

Um, masternodes will always do the mixing. Only now they will do it in quorums.

So it used to be that one masternode would be chosen to mix funds, and the result would be "one round" of mixing. The user would then mix a multiple of times to ensure that their information (ip address or tracking of which funds went where) could be logged and followed by corroborating MNs. The chances that a single malicious entity could own enough MNs to be successful became exponentially more difficult with each round.

With the new quorum system, mixing can happen instantly with many rounds. The MNs are also even more randomly chosen as a group as the group will always consist of MNs that belong to (at the moment) one of 4 age groups. They are first assigned an age group. 1/4 of the oldest MNs (according to when the funds were first placed into the account), then the next oldest, next oldest, and finally newest. From each of these groups, a selection of nodes closest to a hash are selected (based upon that same input hash). So there is a random hash that chooses the group of MNs and a mix of ages in each quorum. Lets say 2 in each age group and 8 in each quorum.

So now in an infinitely impossible to manipulate quorum of 8 nodes, you can be certain that they will do their work securely, or the quorum's work will be rejected. Every block, we have MN-count/8 quorums. These quorums can do a certain amount of work each second. They are quick, as only 8 have to talk to one another as a group. They quickly go through all the transactions sent to them and approve/reject transactions. This is somehow sent to the miners who include everything in the blockchain. If the miner doesn't do it properly, the whole block is rejected. Thus, eliminating the so called 50% attack vector.

I'm guessing here, but I can see the mixing simply going from one of these masternodes to another, ultimately allowing for many rounds in milliseconds. There is no time constraint like waiting for a block. The MNs do all the work, and simply send them to the miners for inclusion.

And as far as transactions go, Evan said every transaction will now be instantX transactions, approved by one of the quorums. Again, miners will have no say as to what is included in the blockchain. If they include something that was not approved, it would be rejected. Only their hash will be used to select the quorums. I don't know which hash, it might be the next one, it might be a random one on the chain?? The latter might make sense as we would have to (at this time) have 400+ hashes to put each MN to work (or select nearest to farthest from a single hash?)

Thus, we can already compete quite well against big money transmitters such as Visa, MC, etc... The capacity is limitless.

Another thing that is being developed is the Distributed Applications Programming Layer, DAPI. This DAPI will hook into the Masternode Network, and people will be able to create applications that can query the network. So, at the most basic level, no wallet needs to download gigabytes of blockchain. The MNs, miners and anyone who wants to keep a full copy of the blockchain, stores a copy, and with thousands of MNs, the blockchain is very well distributed. So, they can serve up the necessary information thus all wallets can act as mobile wallets without worry of having incorrect information due to a central point of failure currently a risk in mobile wallets.

This, or at least the foundation for this should be ready by the TNABC in January, where Evan will be presenting. We should also hear more in about 2 weeks when Evan will be presenting at the laBITconf.