1/ What is handwaving & sloppy thinking is ignoring subtle but important differences between PoW & PoS. Such as the *degree of social consensus required* - in terms of both magnitude & frequency.https://twitter.com/killerstorm/status/1044522585663447040 …
9/ Fraud proofs as initially envisioned by Satoshi turn out to be extremely difficult & might actually be impractical. Intuitively it makes sense, since alerts reintroduce the problem of distributed consensus. Not a trivial problem to solve without PoW.
-
-
10/ You need to be extremely careful in designing alert systems because alerts themselves can become security holes. E.g. alerts can be abused to spam/DDOS nodes, or to create conflicting views of the network.
Show this thread -
11/ The current system, as imperfect as it is, is not all that bad. Worst comes to worst, users can either: (a) verify the source code themselves & build from source, or (b) choose to ignore upgrades
Show this thread -
12/ Sure, most users don’t know how to read code, but what’s most important is economically important nodes like exchanges & wallets. They can & SHOULD verify every protocol changes. They are incentivized to do so because if things fuck up, they would sustain the most damage.
Show this thread -
13/ As for the rest of the users, yes they will have to trust someone for upgrades. But upgrades don’t happen too frequently to begin with, and the trust issue is another reason to strongly advocate for earlier-rather-later protocol ossification.
Show this thread -
14/ > It seems that pro-PoW people just gave up on analyzing the system, they just say "It works because nobody have broken it yet” Another baseless claim. Bitcoin security is a very active area of research. We've merely scratched the surface. Seehttps://twitter.com/hugohanoi/status/1027667455773212673 …
Show this thread -
15/ What PoW proponents are saying is that PoW is the most secure system both in theory & practice, with a 10-year track record. Can PoW fail? Sure. Can PoS succeed where PoW might fail? Absolutely freaking not.
Show this thread
End of conversation
New conversation -
-
-
It is only "extremely difficult" if you require 100% liveness, which is a stupid idea to begin with. If you let the user to decide the level liveness he wants, things are much simpler. For example, some of users can go more paranoid and prefer to halt system even if it's false
-
alarm. If there are too many false alarms they can just disable them. You can design much better pragmatic systems if you assume that users are not stupid immutable robots, but actually know and decide for themselves what they want. There's a well known attack vector that an
-
attacker controlling user's network might feed user's full node invalid chain. Since this cannot be patched in 100% decentralized way, developers prefer to assume that the problem does not exists. I've seen many advice to run a full node, but none of them mentions that one also
-
needs to check that his full node is on the correct chain.
End of conversation
New conversation -
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.
