Charles Hoskinson Drops Cardano 1.3 and Ouroboros / Casper Comparison Debate

Cardano 1.3 Has Been Released to GitHub

Charles Hoskinson, founder of the famous network and cryptocurrency Cardano (ADA), has announced the release of the Cardano 1.3 that can be downloaded from GitHub.

cardano update https://t.co/a1LPKQhrJN

— Charles Hoskinson (@IOHK_Charles) August 6, 2018

Mr. Hoskinson released the information about Cardano on August 9, on his personal Twitter account.

The Cardano SL 1.3.0 features a wallet API v1 for exchanges, a subscription status that has been added to node information, an improved error message for missing charset in API calls and support for sending raw data on the network.

Cardano 1.3 now signed and released to Github: https://t.co/t81CzOZIoQ download and enjoy!

— Charles Hoskinson (@IOHK_Charles) August 9, 2018

Cardano 1.3 Update

Other bug fixes and improvements have been performed:

Faster sending and receiving of blocks:

Sending and receiving blocks on the network works better after this release. Deserialization performance has been improved by optimizing memory usage. Furthermore, the blocks will now be downloaded concurrently without batching.

Optimized block storage:

This is also an important change made because the block storage is now optimized by consolidating block and undo data in a single file. In this way it is possible to reduce disk use and improve performance when reading and writing blocks.

High (and recurrent) I/O traffic in wallet reduced:

“I/O spikes in traffic were caused by large logs being flushed,” the github page reads. “This issue has been fixed.”

Time error between user’s computer and the network:

In this case, the endpoint ( /apo/settings/time/difference) for calculating the time difference between a user’s computer and Cardano network was sometimes returning incorrect values. The situation was not properly handling time needed to request the current time from NTP servers and to get the response.

Ethereum's Casper vs Cardano's Ouroboros

After the release of Cardano 1.3, IOHK took a shot at the Ethereum Network by doing a comparison of the two different Casper and Ouroboros Protocols.

How does Casper compare to Ouroboros? Read the latest blog post by Aggelos Kiayias, chief scientist at IOHK, on the differences between the proposed Ethereum protocols and Cardano’s consensus algorithm: https://t.co/BgLeWpcdbm

— Cardano Community (@Cardano) August 9, 2018

Here is our analysis of what was found in the blog post. Below we have added a Twitter spat between IOHK Charles Hoskinson and Ethereum's Vitalik Buterin.

Main Differences Between Casper and Ouroboros

Ouroboros is an analyzed protocol that features mathematically proven security guarantees that are based on clearly specified assumptions. Additionally, Ouroboros offers stake-based finality with strong guarantees in terms of the amount of stake that backs up the honest operation.

There is no information about Casper and how the protocol operates. Casper seems to provide weaker guarantees in terms of how much stake the adversary needs to control to disrupt the protocol.

It is important to mention that PoS ledger consensus protocols should satisfy persistence and liveness, because it is necessary to ensure the ledger is final and immutable, and that the transactions are broadcasted by honest parties. But these properties cannot be proven to be unconditionally; they will have to rely on certain conditions.

The Ouroboros protocol defines the participant’s programs and the executions and interactions related to them, including their communication. This model allows for mathematically precise security guarantees that are satisfied by any execution. This is very important because the claims about it are concrete and there is nothing left up to interpretation.

Ouroboros provides stake-based finality and it does it protecting against a malicious coalition controlling any amount of the total existing stake in the system.

Blockchain protocols work increasing the reversibility of a block as the number of blocks added on top of it grows. This designs allows these protocols to work in the strongest possible adversarial settings and still be very efficient.

But it may be a little bit difficult to increase the security and at the same time reduce the stake of trust. Blockchain servers are secure, but they are slower than other systems. Near-instant finality have typically different downsides included downgraded security in some cases. But near-instant finality can be built as a service on top of Ouroboros.

The Casper FFG is inspired by the standard BFT consensus protocols. This means that the protocol cannot operate in the ‘sleepy setting’ and ‘dynamic availability.’ This is an important concern in a decentralized setting where the execution is not meant to be left in just a few centralized actors, instead is distributed proportionally among smaller players.

Casper’s white paper reads:

“The inactivity leak introduces the possibility of two conflicting checkpoints being finalized without any validator getting slashed.”

This affects the incentives of those running the protocol. Ouroboros allows for natural and incentive-driven aggregation of stake into stake pools.

Rebuttal to Vitalik: https://t.co/hNhyCLU6Kv

— Charles Hoskinson (@IOHK_Charles) August 9, 2018

Sorry, Charles, this is pathetic.

> Regarding Casper, we are not aware of any currently published source that sufficiently describes the protocol’s mode of operation nor any provable guarantees about ithttps://t.co/SHKWyUOLqVhttps://t.co/LRujTx4hCf

— Vitalik Non-giver of Ether (@VitalikButerin) August 9, 2018

Sorry! Wrong paper: https://t.co/41oRHc5jSy

— Vitalik Non-giver of Ether (@VitalikButerin) August 9, 2018

“Casper requires a ⅔-fraction of deposited stake to be controlled by honest parties. Ouroboros is proven to achieve persistence and liveness under the assumption of honest majority”

This completely misses (i) *incentive* properties, (ii) network synchrony assumptions

— Vitalik Non-giver of Ether (@VitalikButerin) August 9, 2018

Regarding Praos being partially-synchronous, I never got a reply on this: https://t.co/CtmyiL2KJO

I'm ~90% sure that Praos's partially synchronous guarantees are NOT the same thing as safe-under-asynchrony finality.

— Vitalik Non-giver of Ether (@VitalikButerin) August 9, 2018

> Casper makes no explicit claims about the network setting it operates in

This is just a lie. Eg. page 4 of the Casper FFG paper: “Plausible liveness means that, regardless of any previous events (e.g., slashing events, ****delayed blocks****, censorship attacks, etc.)”

— Vitalik Non-giver of Ether (@VitalikButerin) August 9, 2018

The FFG paper defines safety very clearly: two conflicting checkpoints cannot both be finalized unless >= 1/3 are Byzantine. Regardless of network assumptions. And the proof, had you bothered to actually engage with it, clearly does not use network assumptions at all.

— Vitalik Non-giver of Ether (@VitalikButerin) August 9, 2018

The incentive properties of finality, enabled by accountability, are also crucial. Once a Casper block is finalized, you need 1/3 to *sacrifice their deposits* to reinstate it. With chain-based algos, you don't have this property.

— Vitalik Non-giver of Ether (@VitalikButerin) August 9, 2018

Mr. Hoskinson has remained quiet during Burterin's tweet storm. One Twitter user did try and get him to respond:

Please, answer @IOHK_Charles

— Ujjwal Rajput (@ujjwalrajputji) August 9, 2018

We did. He called us pathetic liars.

— Charles Hoskinson (@IOHK_Charles) August 9, 2018

No he didnt. He called the rebuttal pathetic and one statement a lie. You’ve turned it around and combined the two and made it person. You really can be a man child sometimes Charles

— CryptoSelecta (@SelectaCrypto) August 9, 2018

No matter which side of the fence you reside on, both are great for the cryptocurrency world

I did read few threads on Reddit between @iohk_charles & Vitalik , great leaders in this space. Not comfortable conversations indeed. There are always different approaches to different problems/ use cases.

— Praveen Deshpande (@praveenpd) August 9, 2018

Hey @VitalikButerin @IOHK_Charles I love academic discussions even if they get a little heated. The proof is in the pudding though, let's see the rollout and that will let us know for sure.

— Jason Benner (@JasonBenner7) August 9, 2018

Want to follow along with the discussion on Twitter or Reddit? Use the link above or click this reddit post below to add in your own thoughts.