network privacy and mixnet §
research §
- Crypsinous over Mixnet contains the summarization of investigation into how Crypsinous will work in combination with the Mixnet
- In the development section below, we will go into further into details in terms of how this researched affected the specification.
- The discussions are still open and the specification is still prone to change.
development §
- Updated the Mixnet Specification according to the new research and analysis in the Crypsinous over Mixnet document. The gist of the document update is that every node emits real/cover packets with a rate in a Poisson distribution.
- A slot leader publishes 3 block proposals and several drop cover packets within a slot (a slot is published every 20s)
- At that time, all of the other nodes emit only drop cover packets (“decoys”) within a slot.
- The target mean packet delay is 2s.
- Based on this:
testnet §
development §
private PoS §
research §
- Researching on a potential problem of wealth concentration in the token engineering model that was reported by DarkFi since it contradicts the results we have calculated in the Validation Rewards
- Performed additional research in terms of feasibility of learning total stake: the problem here is that we cannot compute relative stake based on what is the Crypsinous paper.
- Refined rigorous lower and upper bounds on the probability of stake greater than some threshold, i.e. “confidence”. The details can be seen in Notes on mathematical and statistical aspects of proof of stake consensus mechanisms
- The upper bound was only available for prob. 0.9 and higher and a new version also allows for lower prob.
- The latter is needed for a more accurate estimation of the number of layers in the mixnet.
- The asymptotic result for the same probability also can be used only for higher values and extending this result for lower values is currently in progress
data availability §
research §
- Writing of the new protocol specification complete - it is still open for discussion - comments and proposals with valid reasoning can still adjust the specification - anyone can comment.
- For the historical records, the trail of thoughts, scenarios and improvements can be seen here.
- The new specification was also looked at by the Nescience team - it has no cryptographical weaknesses.
- Implementation expected to start at the beginning of January.
- In the future, we will evaluate additional Data Availability structures (based on internal literature) to see how they compare to what we have.
- the DA node read/write API implementation is in progress - we have reviewed and reevaluated the initial plan to implement DA data dissemination and retrieval flows and created an action plan based on that.
development §
- We have performed simulations to confirm there will be no “flatness” issues with Carnot implementation.
- We are currently in the process of finalizing the Carnot paper - the simulations action plan will also be shown there.
- Found additional proofs that simulations and rust Carnot implementation acts as expected in varying committee overlays - more info on Discord.