Hacker News Reader

Comment by redka | Hacker News Reader

IOTA actually seems pretty revolutionary. I don't think it's a finished product yet but if you want some info then it's a DAG ("tangle")[1] instead of a blockchain and offers 0 fees and possibly huge scalability. There are some products already using it: RuuviTag[2], Bosh XDK Iot[3], PoC ElaadNL charging station[4], Modum[5]. Once they prove they can go without the "Coordinator"[6] I'm going to be completely sold on the project.

[1] https://iota.org/IOTA_Whitepaper.pdf

[2] https://lab.ruuvi.com/iota/

[3] https://xdk.bosch-connectivity.com/

[4] https://medium.com/@harmvandenbrink/how-elaadnl-built-a-poc-...

[5] https://www.reddit.com/r/Iota/comments/6vl57m/iota_modum_ask...

[6] https://domschiener.gitbooks.io/iota-guide/content/chapter1/...

View on HN

The principles could be relevant (although I haven't verified (or read an independent review of) the proofs).

However, at least the execution is dubious: not only was a major cryptographic vulnerability found a few months ago[0], it highlighted that

1. they made their own unproven crypto hash, which is a red flag, (in production, you should always prefer crypto that has survived many years of cryptanalysis) and

2. they were really not up-to-date on cryptanalysis techniques, as being vulnerable to differential cryptanalysis is, as Bruce Schneier puts it, "a rookie mistake".

Worse than that, they seem to rely on security through obscurity, purposefully making it hard to analyse their systems by making it base 3 (instead of the obvious base 2 that all of practical computer science relies on).

Finally, their justification for the coordinator is also a red flag[1].

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

[1] https://medium.com/@ercwl/iota-is-centralized-6289246e7b4d

The coordinator will be open-sourced and in all seriousness, even bitcoin had a sort of coordinator in the beginning, people aren't even forced to use it/can bypass it, but it's recommended to use it for now. So your statement isn't 100% valid.

Regarding crypto, you can read their response & reasoning here: https://blog.iota.org/curl-disclosure-beyond-the-headline-18...

"The replacement Kerl hash function is unmodified KECCAK-384 that only converts its input and output from/to 243 trits to 48 bytes using basic two’s complement. KECCAK-384 is well vetted and researched."

1. It doesn’t matter whether the software that implements the central coordinator is open source, because the entire challenge of creating digital money is decentralized trust. Chaumian Cash solved digital money with a central issuer long ago.

2. Bitcoin had no privileged nodes, nor does it now. All miners compete on the same terms. A central coordinator by definiton does not, or it wouldn’t be centralized.

Just to clarify, since I misread your quote and think others might as well:

> "The replacement Kerl hash function is unmodified KECCAK-384 that only converts its input and output from/to 243 trits to 48 bytes using basic two’s complement. KECCAK-384 is well vetted and researched."

"Replacement" really does mean replacement, i.e. what they substituted in after the researchers successfully attacked their original hash function. (The new one is called "Kerl" and the old one is called "Curl", which seems confusing.)

I'm not sure how asserting that the replacement is well-researched addresses the OP's point that their history of (a) rolling their own hash function, and (b) not adequately considering differential cryptanalysis, is a red flag. True, that specific vulnerability can now be considered fixed, which might not be clear from the OP's post - but it's still evidence of incompetence, which doesn't bode well for the quality of the rest of their design.

This is a common misunderstanding. IOTA never deployed a vulnerable hashfunction. They had precautionary measures in place and thus had Curl there to test it out, which worked out brilliantly. Keep in mind that IOTA asked the team to attack Curl, not the other way around. This was planned.

Curl is meant to be a lightweight crypto for IOT, a field of very active research. None of this is controversial to anyone that isn't looking for things to latch negativity onto.

> Keep in mind that IOTA asked the team to attack Curl, not the other way around. This was planned.

This seems to contradict the researcher's own post [1]:

> We discovered a vulnerability in IOTA after reviewing their code on GitHub in July. We disclosed what we found to the IOTA team on July 14th, and have been in contact with them since then as we discovered new issues and exploits.

Finally, even if Curl is meant as a new, lightweight hash function, it was broken by differential cryptanalysis, not some novel, exotic attack vector. Sounds like it needs a lot of work before it's fit for purpose.

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

Yes, Ethan was then forced to admit that the IOTA team actually approached him in May.

Can you provide a link please? All I see in the comments to the researcher's piece is an IOTA advisor threatening a libel suit - a really good sign that they "really care" about their technical issues.

I highly suggest you read

https://blog.iota.org/curl-disclosure-beyond-the-headline-18...

Ok. I've read it. Nowhere does it mention IOTA contacting the researchers in question in May.

This article also answers the wrong question. If the crytocurrency is not cryptographically secure all that stands between an attacker and a victim is a piece of malware or social engineering. The fact that the researchers didn't go all the way and document a specific attack that could be performed tomorrow does not mean that Curl was secure in practice.

Finally this continues to fail to address many salient points. Like why use trits? Why wasn't kekkac used from day one?

IOTA, and others, were not analyzed from adversarial known/expert researchers yet. Period. For more serious DAG oriented research you can check SPECTRE and the papers in this list [1].

[1] https://docs.google.com/document/d/1J8hehbnZWzcIUMQcxMiGbjz8...

> […] offers 0 fees […]

I’m highly skeptical of this claim since there’s a cost to running any system. It seems more likely that they just made fees implicit rather than explicit (as in Bitcoin) but, regardless, someone needs to be paid to keep servers running.

In bitcoin you don't get paid to run full nodes. Only the miners get paid for verifying transactions. Verifying transactions is done by the users themselves with iota, so there is no need for miners and no need for fees.

Their claim of 0 fees will not last. It means DDOSing will be trivial, and their network will be crippled.

You have to do PoW when submitting a transaction. So DDOSing will not be trivial.

But the PoW is for small IOT devices right? Or at any case it is for regular users, and does not take them minutes to compute. I cannot see how that is resistant to an adversary with dedicated hardware.

> You have to do PoW when submitting a transaction.

Then there is a price; it's just measured in watt-hours of energy spent instead of denominated in cryptocurrency.

> Verifying transactions is done by the users themselves with iota [...]

Unless the cost of verifying transactions is zero, this just means that the cost is now measured in CPU cycles at the client. In the end, this may be higher than whatever you pay with Bitcoin, depending on how much work you need to do. This is what I mean by "making fees implicit" - there's still work to be done, but not putting a price on this work doesn't mean it's free.

Reply on news.ycombinator.com