Hacker News Reader

Comment by xwvvvvwx | Hacker News Reader

Not really deep into this, and haven't even read the whitepaper, but looking from the outside there are a lot of red flags around this project:

- Use of ternary logic [1]

- Writing their own hash function [1]

- Claiming that the flaws in the hash function were 'copy protection' [2]

- The Github issues page for their wallet client [3]

I've heard that the DAG approach has merit, but from what I've seen I would not trust this team to execute on it.

[EDIT]: Almost forgot the black-box closed source 'central coordinator' [4]

[1]: https://medium.com/@neha/cryptographic-vulnerabilities-in-io...

[2]: https://gist.githubusercontent.com/Come-from-Beyond/a84ab861...

[3]: https://github.com/iotaledger/wallet/issues

[4]: https://www.reddit.com/r/Iota/comments/7c3qu8/coordinator_ex...

View on HN

1: By using a ternary number system, the amount of devices and cycles can be reduced significantly. In contrast to two-state devices, multistate devices provide better radix economy with the option for further scaling

2: Why is explained here: https://blog.iota.org/the-transparency-compendium-26aa5bb8e2... - CyberCrypt is hired to review and audit it.

3: The wallet is secure and does what it needs to do, no, it's not very pretty or user friendly but it works. A new wallet (Trinity wallet) will be released very soon.

> 1: By using a ternary number system, the amount of devices and cycles can be reduced significantly.

How does that work? Trying to find information, it looks like it's designed for devices of the future that don't exist; so it's built on assumptions that may be plain wrong.

It would be ideal on trinary hardware, but it can work with a trinary emulator for now. I expect to see a Trinary chip very soon. CfB, one of the core devs of IOTA, founder of NXT and inventor of PoS has a team that has been working on Jinn (trinary cpu's) for the last 6 years.

So they're betting that it's the hardware of the future. And not something like the Mill CPU, which has been in development for much longer and it seems it will be able to execute all the millions pieces of existing software, faster and with less energy.

ad 1. Citation needed

ad 2. From the article: "Let’s begin with the common sense, IOTA’s Tangle is a significant leap forward from blockchain". Why do you believe this subjective superlative when the protocol is still in beta and heavily contested? It's like saying: "Fusion reactors are a significant leap foward", while no production ready fusion reactors exist. You cannot make this claim unless you're willing to attest to the correctness and production-readiness. This seems common sense to me.

ad 2: What is this document? "The Transparency Compendium", "Transparency Report", "Blog post"? All of them are mentioned, but there is no mention whether this is an official statement, or just a sputter of random thoughts by the author. It seems official, but isn't.

ad 2: No mention of CyberCrypt in that article. Extraordinary claims require extraordinary evidence. Also, Occam's Razor. That's what I like about the original bitcoin protocol, it's as simple as necessary, but not simpler.

ad 3: It doesn't need to be pretty or user friendly. The complaints are about people losing their IOTA's and all kind of other complaints. Please address these complaints head-on, instead of just making blanket remarks about user-friendliness.

Hi,

-Can you explain why the use of ternary logic is red flag? I can explain why it is not but you have to justify your statement first. You calling the use of ternary logic a red flag must be based off of something I assume.

-They used a custom Hash functio called "Curl". But they now replaced it with KECCAK-384. Curl is now being reviewed by a third party.

-For this, I will let CFB (one of the devs) to explain. https://medium.com/@comefrombeyond/cfbs-comments-on-https-ww...

-wallet is a nuisance for some users I agree. But majority of users are fine with the wallet. Of course you can only hear those who had problems.

Ternary logic is a pretty niche thing. As I understand it the claimed benefit is improved efficiency / elegance for some mathematical operations when run on a ternary computer.

However since all modern hardware is binary, the use of ternary logic only serves to introduce an unnecessary software based translation layer, simultaneously increasing complexity and reducing performance.

I would consider this to be an example of very poor engineering judgement. Good software is simple software. I would be extremely concerned if a colleague suggested such needless overcomplication.

However since all modern hardware is binary

They're actually working on trinary hardware with the closely related JINN project.

https://bitcointalk.org/index.php?topic=1273040.msg13159701#...

How come the links to jinn repos don't work? https://github.com/JinnLabs/

I would not consider in poor engineering judgement. It is a judgement yes. By using binary, you don't need conversion but lose computational power. By using ternary, you lose thru conversion but gain computational power.

In the end of the day, you have to choose as a dev.

So to your point, I don't consider it a red flag. You can even interpret it as "thinking outside the box". Yes ternary is more complex, that's what the devs are here for.

> By using ternary, you lose thru conversion but gain computational power.

How does that work? Isn't it being emulated on binary hardware?

Using ternary logic causes a lot of things (like the custom cypto) to be non-standard and thus hard to evaluate. "Rolling your own crypto" is a red flag in itself, but making such choices while everyone else in the blockchain/DAG world mainly uses tried-and-tested building blocks, seems like a red flag.

And in the end they are "just" red flags (= warning signs), which might hint at fatal flaws, but also might not.

https://eprint.iacr.org/2016/1159.pdf

The above is a proper paper on how to do math for a DAG based protocol.

Serguei Popov, the research mathematician behind IOTA DAG based protocol, claims that it is impossible to make proofs behind the tech. [1]

[1]: https://medium.com/@johndom/iota-and-spectre-64ee12d9b1a8

Reply on news.ycombinator.com