Quote from: wiebemarten on November 13, 2017, 12:09:17 PM

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Recently I read the whitepaper about Ripple's consensus mechanism.
Of course, it has slightly other invariants than Bitcoin. For instance, in Ripple, when less than 80% of the network is honest, no honest transactions can be sent so the network grinds to a halt. On the other hand, if less than 20% of the nodes is honest, malicious transactions could be stored permanently. In Bitcoin, 51% is enough to rule the network and there is not a lot of introspection to decide whether the majority is still honest.
Another 'problem' that Ripple has, is that all users need to have a good 'Unique Node List'. If the UNL is not a proper representation of the network as a whole, then it might be possible to game the system earlier.
On the other hand, because being thrown off of the UNL by many nodes is a very valid response to misbehaving that is very bad for the node that tried to cheat (you are now effectively excluded from the network), this seems to keep all nodes 'in check'.

I saw someone describe this this as 'Ripple allows you to shoot yourself in your foot. But it also allows you to react when someone tries to shoot you in the head'.

At least for me personally, Ripple, at least for financial transactions, seems to have the better method to go from transaction requests to an ordered transactions list.
Keep in mind though that I have not worked with Ripple directly myself, so I am not yet aware of the warts that it might have. (Usually warts of any language, framework, technology or tool are harder to find, because they are not as widely publicized and talked about).

Proof of Work is definitely an 'arms race' with no end and we need a replacement. I think we can all agree about that. It does not really matter for the security of the network if 100 nodes mine or 10000 nodes mine, but that the latter takes a whole lot more energy.

I believe there is a lot of research focused in this area right now. Besides Ripple, here is a list of other tech I happen to know of that is being worked on, including some notes about their current progress:

- - Proof of Stake: Still not mathematically proven enough to be acceptable. (Do note that there are multiple different variants of Proof of Stake that each work slightly differently)
- - Proof of Elapsed Time: Greatly hyped by HyperLedger, but it is proprietary technology by Intel that only works if you have a certain Intel chip, and places trust in the Intel company. I don't know enough about how this should work to know why it wouldn't be possible to e.g. make a fake hardware chip.
- - Proof of Burn. Not entirely sure how the progress is on this one.
- - The IOTA Tangle. Seems a very promising method to build a DAG rather than a chain. A problem of which I don't know how they want to solve it, is how it is prevented that new info is added as children of 'old' elements, essentially 'inserting something in the past'.
- - The Hashgraph. Currently closed source (the team does claim that it will be published as FOOS at some point though). Also builds a DAG in a different way than the Tangle does. It is easier to see how the network as a whole prevents 'insertions in the past'. What is not yet known, is how/or even if the Hashgraph can work in a public context; it currently requires a static number of nodes (where all nodes know each other at the start) and where a supermajority of nodes is online at all times. There are some vaguely-described extensions to this, but these do not explain how someone could join and leave at any time as people can do with other ledgers.

This list is by no means complete.

~Wiebe-Marten
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1
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=4i5F
-----END PGP SIGNATURE-----

More on Hashgraph:
Based on Hashgraph concensus. The creators claimed it is 50,000 faster than Bitcoin Blockchain and would also allow Smart Phones to act as Nodes.
Article on Hashgraph:
https://breathepublication.com/blockchain-just-became-obsolete-the-future-is-hashgraph-de4948609cbf