SoylentNews is people
SoylentNews
Submitted via IRC for SoyCow8317
Bitcoin isn't the only cryptocurrency on a hot streak—plenty of alternative currencies have enjoyed rallies alongside the Epic Bitcoin Bull Run of 2017. One of the most intriguing examples is also among the most obscure in the cryptocurrency world. Called IOTA, it has jumped in total value from just over $4 billion to more than $10 billion in a little over two weeks. But that isn't what makes it interesting. What makes it interesting is that it isn't based on a blockchain at all; it's something else entirely.
[...] Instead of a blockchain, IOTA uses a "tangle," which is based on a mathematical concept called a directed acyclic graph. Sønstebø says his team pursued an alternative approach after deciding that blockchains are too costly—it has recently cost as much as $20 per Bitcoin transaction because of high demand—and inefficient to operate at the scale required for the Internet of things.
[...] So IOTA has dispensed with the miners. Instead, when a user issues a transaction, that individual also validates two randomly selected previous transactions, each of which refer to two other previous transactions, and so on. As new transactions mount, a "tangled web of confirmation" grows, says Sønstebø.
(Score: 0, Interesting) by Anonymous Coward on Sunday December 17 2017, @09:16AM (10 children)
Inefficient for the IoT, sure that might be the case, but I wonder what exactly is inefficient? The chain size I guess, but where does IOTA keep it's transactions? That too will grow. Bitcoin transaction costs aren't so high due to the fact that bitcoin uses a blockchain, strange that a dev uses that false argument, he should know better if he is working on these type of things.
The concept is interesting, but I wonder how the things resolves around double spending if the network splits. How it is written down above it does not seem very secure if parts of the network split or very few transactions happen.
(Score: 2) by tonyPick on Sunday December 17 2017, @10:47AM (4 children)
There's more technical information here, which covers some of these details....
https://blog.iota.org/a-primer-on-iota-with-presentation-e0a6eb2cc621 [iota.org]
(Score: 0) by Anonymous Coward on Sunday December 17 2017, @11:26AM (3 children)
Yeah, they mention the network split as a feature, but if I would spend in both parts of the network and reconnect later (when transactions have been confirmed in both), how will that be resolved? I can't find this back in the article.
(Score: 5, Informative) by tonyPick on Sunday December 17 2017, @12:30PM (2 children)
(disclaimer - not an expert on this, so I may be wrong, but....)
I think the answer is that the attack you're talking about is called a "parasite chain" case: It's covered as a deliberate attack in 4.1 in this whitepaper https://iota.org/IOTA_Whitepaper.pdf [iota.org]
and here: https://forum.iota.org/t/iota-double-spending-masterclass/1311 [iota.org]
Basically the chain can be inconsistent for short periods, but when the parasite chain (i.e. the bad doublespend chain) tries to merge back it will get rejections and fall behind the main chain and eventually die out. Merchants can (apparently) decide when they want to accept payments in terms of verification from the transaction back to a confirmed transaction on the main tangle, so they can avoid this.
In a "offline" case where the trades between nodes are done away from the main tangle then the transactions won't be confirmed until the offline tangle merges back with the main tangle - so if you run in an offline group in theory that can fall into inconsistency with the main tangle, but you'd be aware of that case and wait to merge back to the main tangle to actually confirm state.
(all AIUI - any experts want to chime in?)
(Score: 1, Interesting) by Anonymous Coward on Sunday December 17 2017, @02:51PM
But given the limited 'tangle', it seems like it wouldn't require a huge number of nodes to fake tangles on transactions, basically upping your chances of successfully gaining new currency by figuring out easy to process checksum chains, having enough nodes to not trigger whatever 'spam processing' guards are in place, and then basically have all your nodes 'randomly' process the same chains of transactions, possibly changing identification addresses periodically so the transactions can't be flagged as verifiably false.
This would not be a cheap or easy attack to pull off, but like the 51 percent attack on bitcoin it is entirely feasible if enough interest in the coin, and money in the mining or speculation of the currency takes off.
(Score: 3, Insightful) by maxwell demon on Sunday December 17 2017, @09:31PM
So what happens with all the payments that had the misfortune of selecting a tip from the "parasite chain"? Are they all invalidated?
In that case someone could "poison the well" with an intentional large number of double spends, invalidating a large number of unrelated spends. Even if the double spends get invalidated, there's a big damage through the collaterally invalidated transactions. That might be an intentional attack in order to push down the prices (by reducing trust in the currency).
The Tao of math: The numbers you can count are not the real numbers.
(Score: 5, Informative) by tonyPick on Sunday December 17 2017, @11:06AM (3 children)
And, coincidentally, just saw this much more critical review, on the frontpage of hackernews...
http://codesuppository.blogspot.com/2017/12/iota-tangled-mess.html [blogspot.com]
https://news.ycombinator.com/item?id=15944112 [ycombinator.com]
(Score: 2) by FatPhil on Sunday December 17 2017, @10:44PM (2 children)
I got lead to this:
https://medium.com/@neha/cryptographic-vulnerabilities-in-iota-9a6a9ddc4367
which, in addition to the "balanced ternary is more efficient" claim *bullshit* (because nobody's developed a way of storing a trit more efficiently than two bits, even those russkies who had the idea half a psychadelic-infested century ago) is enough to make me consider the particular implementation of a tangle architecture a trainwreck. Sure, use the DAG idea elsewhere, but in a context that isn't otherwise flawed or simply dumbarsed.
Great minds discuss ideas; average minds discuss events; small minds discuss people; the smallest discuss themselves
(Score: 2) by maxwell demon on Monday December 18 2017, @07:27AM (1 child)
From the linked blog entry, emphasis by me:
Wait … what? There are people who expect something Microsoft is involved in to not have serious security vulnerabilities?
The Tao of math: The numbers you can count are not the real numbers.
(Score: 2) by FatPhil on Monday December 18 2017, @09:37AM
Great minds discuss ideas; average minds discuss events; small minds discuss people; the smallest discuss themselves
(Score: 2, Interesting) by dc on Sunday December 17 2017, @04:36PM
"...what exactly is inefficient?" You could point to several things.
Bitcoin's proof of work is a cryptographic puzzle that must be solved, in addition to calculating the debits and credits that happen with each transaction applied to a block. The system adjusts the difficulty of the puzzle as more power (computational and electrical) is participating. So more resources don't speed up the system or allow more transactions. More resources means more work, not more output. (One could argue this additional work means the system is more secure, however.)
It's inefficient in another way that I find more interesting. Let's say you and I are miners, and we both solve the cryptographic puzzle at nearly the same time. We've both done perfectly valid calculations. There's nothing "wrong" with either one. Yet bitcoin system will eventually accept one of these blocks, but not the other. Here it is inefficient in that perfectly good work goes discarded.