Hi Shunsai,
I was very sleepy as I was writing that and went off to sleep with numerous typos and incomplete logic expressed in haste. Immediately upon waking (and actually before reading the new comments because I hadn’t reloaded the Steemit web page), I edited my analysis of Obelisk and I think you will find my current analysis to be more in agreement with your assessment after you correct your misunderstanding below. Obelisk is intriguing for me because it has the potential to augment and improve extant non-proof-of-work systems. As you can see by @life.thoughts’ comment, apparently Skycoin also has a new unpublished algorithm in which they may have made an analogous insight as I have (and I haven’t published all my thoughts). And notice they’re not publishing it in open source until they attain first-mover-advantage (FMA), which is what I’ve also been telling you I should do as well.
Also it is very important to incorporate my thoughts about the value and nature of security.
Note I may be attempting to contact Synth again. I had tried to contact him in 2015 on Bitmessage but failed. There’s another synergy in that they’re using Go and I want to create a transpiler to Go to add polymorphism abstraction. Also Synth has been rambling (as user @skycoin) in the bitcointalk.org thread since 2015 about the similar problems with trust in our society. So he and I have a similar ideological stance. However, there may exist some monkey business and internal discord in their team as recently reported.
WP2's assumption (0.3.2 page 2) that […]
I already stated in my blog that whitepaper is junk. Let’s not discuss it.
The entire analysis (in WP1) is done using mean field theory perhaps under the assumption that the mean behavior of the nodes fully defines the desired (and undesired) properties of the system. Unfortunately, it is the edge cases, not the mean behavior, that are the most critical to a system like this.
The MFT model is characteristic of the adversarial power needed to corrupt and create edge cases. AFAICT, we wouldn’t gain any attack model insight by modeling each node individual.
And even without using the analysis for all the edge cases, all the protocol can offer is a mere 13% failure tolerance.
I already mentioned that as a significant disadvantage in my analysis (even before I edited it). However note that it requires the non-Byzantine nodes to inadvertently choose to allow an adversary into their web-of-trust. And this is a significant difference from Stellar SCP which presumes that they must allow the adversary into their web-of-trust if that adversary has been allowed into a significant portion of the web-of-trust of other nodes.
Note that it is 13% of nodes not stake. Also note that the nodes are rather inexpensive to create (at a cost of approximately $70/hour per 10,000 nodes).
You’re correct about the 13% but incorrect to presume the adversary can just add nodes. The adversary must be able to induce other nodes to accept his Byzantine nodes into the web-of-trust choices of other nodes. The whitepaper explains this in Fig. 5 of §7.1 Sybil Attack on pg. 11.
I could be wrong but I am completely missing the point or usefulness of this protocol.
Yeah you missed the point that non-Byzantine nodes make their choices of which other nodes they trust.
However, please note I have added a Disclaimer section for Skycoin to the blog. I’m not convinced it isn’t a scam.