4 ms·
There's a proof in Bitcoin's paper that shows the probability of someone catching up with the main chain. That's actual real data that we can measure. Of course
by sktrdie 11y ago
There's a proof in Bitcoin's paper that shows the probability of someone catching up with the main chain. That's actual real data that we can measure. Of course nothing's 100% true in the real-world, but with PoW you get very real numbers on the probability of someone catching up, which decreases drastically after each confirmation.
On the other hand with PoS I still haven't seen any data whatsoever. I asked a simple question: since there's no cost for someone to simulate a chain, what's stopping peers from doing this? I know there are checkpoints and things like Nxt don't allow deep reorgs, but for new users joining the network it means you need to point them to trusted parties that can give them this information: they can't trust the chain independently.
- kushti 11y agoIn the first place, by using modern ASIC it's possible to draw a chain from the genesis having no difficulty changes(by setting proper timestamps), and it will be longer than network's probably as latter has delays more than 10 mins pretty often. So Bitcoin has checkpoint as well ( https://github.com/bitcoin/bitcoin/blob/86cfd23f68367af072500b1758a4c446cdd36e74/src/chainparams.cpp https://github.com/bitcoin/bitcoin/blob/86cfd23f68367af07250... ). In the second place a chance of drawing a long chain suffix with cumulative difficulty better than network's is negligible. Short-range attacks are possible, but they are not a huge problem. Please read the forum post with some results got with simulations https://bitcointalk.org/index.php?topic=897488 https://bitcointalk.org/index.php?topic=897488
- sktrdie 11y agoWhat do you mean? If I look at the latest block hash and difficulty I can actually measure how much hashing power went into finding such block (and it's a lot). How would you simulate that with timestamps?
- kushti 11y agoYou can generate blockchain since genesis setting timestamps every 10 mins so no any difficulty change. And that chain drawn to the moment will be longer than network's.
- sktrdie 11y ago@kushti I think you might have misunderstood how the consensus protocol works in Bitcoin. It's not the longest chain, but the one with the most work put into it that is accepted.
- kushti 11y ago@sktrdie Can you please send me link to a spec / source piece proving that?
- sktrdie 11y ago@kushti sure: http://bitcoin.stackexchange.com/questions/936/how-does-a-client-decide-which-is-the-longest-block-chain-if-there-is-a-fork http://bitcoin.stackexchange.com/questions/936/how-does-a-cl... http://bitcoin.stackexchange.com/questions/29742/strongest-vs-longest-chain-and-orphaned-blocks http://bitcoin.stackexchange.com/questions/29742/strongest-v... Pieter Wuille is a core developer afaik: "The actual rule always favors the chains which required the most "work"."
- kushti 11y ago@sktrdie Thanks for the links! Ok it seems at the moment Bitcoin uses some kind of a cumulative difficulty. P.S. Joining a network is implemented in the same way usually for both PoW/PoS currencies.
- sktrdie 11y agoBy the way, if you're aware of any peer-reviewed papers out there that explain how PoS works, I'd be interested in reading them. Academia has been quite shy on the subject.
- kushti 11y agoUnfortunately, I dont' know any peer-reviewed proof of PoS (im)-possibility, threat model etc. Time is needed here probably, we got good formal framework of PoW only in 2014, 5 years after Nakamotos' WP(I mean Backbone's paper https://eprint.iacr.org/2014/765.pdf https://eprint.iacr.org/2014/765.pdf).
- tromp 11y agoyou can simulate a longer genesis originating chain with constant low difficulty by keeping block intervals at exactly 10mins, but the consensus rule in Bitcoin is not to pick the longest chain but one with the largest cumulative difficulty. The Bitcoin checkpoints are rather redundant and can be phased out with no loss of security.