ENG

What I don't understand is this: why the fuck would any blockchain use ENG as a second layer? What's the incentive to make this happen?

"This makes is possible"

Doesn't even check.

its almost as if he was holding bags
hmmmm

Scalability and Privacy.

reddit.com/r/enigmacatalyst/comments/7rb7ix/enigma_privacy_for_every_blockchain_get_started/

Maybe its a direct quote haha

"The requirement of trust on the participants is also an onerous one; note that, as is the case with many other applications, the participants have the ability to save the data and then collude to uncover at any future point in history. Additionally, it is impossible to tell that they have done this, and so it is impossible to incentivize the participants to maintain the system’s privacy; for this reason, secure multi-party computation is arguably much more suited to private blockchains, where incentives can come from outside the protocol, than public chains."

any counter to this fud?

answer me you zealous brainlets

The scalability part I'm not too sold on, yes it'll be nice to have tx/s done off chain, but you'll have software and hardware solutions like EOS or HPB that will do the job just as well, if not better for different blockchains that require 100's of thousands or millions of tx/s.

Bought a stack of ENG because they can split computation and transaction of data through multiple nodes, with each node processing only parts of the initial data, without having the data being compromised to any of the nodes that make the computation.

Incentive comes at adopting the protocol when the blockchain needs to deal with sensitive data, say medical records; with relative ease.

RLC/ENG/LINK marine reporting in.
After the flippening, all eth transactions will be ran through these protocols.

Not enigma fud at all.

From Vitalik Buterins Jan 2016 article.
blog.ethereum.org/2016/01/15/privacy-on-the-blockchain/

“Secret contracts are like smart contracts, in that they allow executing contracts (i.e., code) with high degree of integrity, which in simple words mean that no one can manipulate the results. However, unlike smart contracts, the underlying input data/state being processed is not visible to anyone, including the nodes in the Enigma network that actually carry out the computation.

Think of a smart contract that lives on the blockchain and tries to find correlations between specific genes and certain diseases. On any blockchain today, this equates to people sharing their genomic data openly for everyone to see. With secret contracts, their data would remain private, because it’s always encrypted.

The way these work under the hood is by one of several means - secure hardware, secure multiparty computation (MPC), or fully homomorphic Encryption (FHE). Enigma employs the first two to enable this with scale (FHE, while theoretically possible, is not likely going to be practical anytime soon).

Note that ZKP/zkSnarks (which are an amazing scientific achievement) are a complementary, but not a sufficient, solution to the problem of privacy. ZKP, in general, allow proving the correctness of a computation to others without revealing the underlying data - but the prover still needs full access to the data. This means that with ZKP someone has to see the data - so you can’t, for example accomplish the example above. There was also an exploration of such a system before called Hawk, which had to rely on a trusted manager with the privacy of the data for this exact reason.”

Source:
medium.com/@EnigmaMPC/computing-over-encrypted-data-d36621458447

Agreed that it is one of the flaws of having a decentralized masternode system where the system's privacy could be compromised through collusion, within the protocol. Say a whale accumulates a vast amount of ENG, and runs enough masternodes that risks the data breach.

But the founder did a recent AMA where he confirmed the amount of ENG required to run a masternode would be a dynamic system, where it can be changed. So theoretically to combat said whale with malicious intents, the team could lower the number of tokens required to be a MN, and further decentralize the system's privacy security by vastly increasing the numbers of participants that could run MN's.

Nuff said

Or I guess the data is encrypted for every node the processes parts of the original data. But is this confirmed unhackable? There's some FUD regarding quantum computing and encryption right?

they would need to control a majority of the network. there would be no reason for all miners not to collude, because there would be no way to prove that they didn't. that's the problem.

however, it's possible that miners would refrain from collusion because ultimately it would cause the enigma network to fail in the long-run. but generally counting on long-sightedness in crypto incentives is a bad idea. my concern is that relying on this sort of game theory would be too risky for dapps that need to use very sensitive personal data.

doesn't address the problem in my post. this is just saying smpc works without considering collusion.

wouldn't*

Encryption of data to single nodes should be able to prevent this, rereading whitepaper now for this part, you got me curious myself.

Yeah that would require a lot of 'Bad' masternodes, discussed in the AMA as you mentioned.

Also Sandy pentland discussing here. (22m mark)
youtube.com/watch?v=1TGfZ55dOb4&t=22m20s

>doesn't address the problem in my post. this is just saying smpc works without considering collusion.
the underlying input data/state being processed is not visible to anyone, including the nodes in the Enigma network that actually carry out the computation.

Incredibly unlikely over 51% of the nodes will be bad actors.
youtube.com/watch?v=bhE0T2tLaHE&t=01m10s

Page 4 of whitepaper dictates a minimal amount of parties required to decrypt the data. So the dynamic number of ENG required to be a MN could temporarily weed out the malicious attempts for a collusion attempts. Pic related.

Page 12 discusses security deposits with a penalty system of losing ENG/BTC to host masternodes to reduce attack attempts on MPC protocols.

Also on top of increasing MN's, they could also increase the t variable, thus by increasing the minimum required parties are required to decrypt the data, it should in theory, prevent collusion attempts.

But this question raises another question, how will they modify the t variable? Is it going to be voted in by the MN's as is with the dynamic system of requiring x amount of ENG to be a MN?

If a 51% attack already did occur, then the democratic approach will fail here. Hmmm.

i see. makes some sense. i think vitalik's concern was that nodes could save their portion of the data, so it would be impossible to determine that an attack had occurred.