The below is an off-site archive of all tweets posted by @lopp ever

September 30th, 2016

@angela_walch Extremely hard, quickly asymptotically approaching economically infeasible. :-)

via Twitter Web Client in reply to angela_walch

@angela_walch IDK, it’s much more convenient to fit “immutable” into a tweet than it is “provably thermodynamically expensive to revert.”

via Twitter Web Client in reply to angela_walch

@gubatron @docbtc But yeah, from some perspectives mutability is a feature. From others immutability is a feature.

via Twitter Web Client in reply to lopp

@gubatron @docbtc Goes back to @aantonop’s first point: nothing is perfectly immutable. The question is what system costs most to mutate.

via Twitter Web Client in reply to gubatron

@gubatron @docbtc If consensus is federated then it should also be trivial cost to change the implementation.

via Twitter Web Client in reply to gubatron

@wsculley @philfrancis77 @aantonop It seems reasonably possible, though we’ll just have to wait and see…

via Twitter Web Client in reply to wsculley

@wsculley @philfrancis77 @aantonop Decentralization is a continuum, not a boolean. I wrote about mining here: https://t.co/k0k3QIP9Ot

via Twitter Web Client

@wsculley @philfrancis77 @aantonop To my knowledge, all long-range attacks are theoretical. I’m not aware of any ever being carried out.

via Twitter Web Client in reply to wsculley

@twobitidiot It really depends upon how much you value immutability and censorship resistance. Avg person may not care, but @wikileaks does.

via Twitter Web Client in reply to twobitidiot

@wsculley @philfrancis77 @aantonop The result is that long-range mining attacks that rewrite history are much cheaper in PoS.

via Twitter Web Client in reply to wsculley

@mzdr0 @paul_btc @ofnumbers @ArthurB @aantonop’s definition is great; now we just need consensus upon it. Perhaps a Twitter poll?

via Twitter Web Client in reply to mzdr0

@paul_btc @ofnumbers @ArthurB @aantonop “Immutability” is a trigger word since people define it differently - nothing is 100% immutable :-)

via Twitter Web Client in reply to paul_btc

@ofnumbers @ArthurB @aantonop You mean because given sufficient hashing power and time, a chain can be rewritten?

via Twitter Web Client in reply to ofnumbers

@ofnumbers @ArthurB It sounds like you’re talking about tamper evidence. Recommend watching @aantonop describe difference in his talk.

via Twitter Web Client in reply to ofnumbers

@ArthurB @aantonop Not if they are censoring transactions that they are legally allowed (or even obligated) to censor.

via Twitter Web Client in reply to ArthurB

Bitcoin Unlimited is accepting proposals to fund projects that research and improve Bitcoin scaling. https://t.co/mO7VZrhwfi

via Twitter Web Client

@ArthurB That’s tamper evidence. @aantonop is referring to cost for miners to collude against users.

via Twitter Web Client in reply to ArthurB

@stanmarion @aantonop Recommend watching https://t.co/z7N64ZV0sa - it becomes prohibitively expensive to maintain a long-term rewrite.

via Twitter Web Client

“Distributed ledgers run by banking consortiums via federated consensus have a cost of 0 to rewrite history. Not immutable.” -@aantonop

via Twitter Web Client

Our ancestors said “it’s as good as written in stone.”
Our grandchildren will say “it’s as good as written in the blockchain.”
- @aantonop

via Twitter Web Client

@stanmarion @aantonop We’re talking long-term past, not short term. It quickly become uneconomical to change past the further back you go.

via Twitter Web Client in reply to stanmarion

“Proof of work is also Proof of Stake. But Proof of Stake is not Proof of Work. With PoW, miners are staking energy (money.)” -@aantonop

via Twitter Web Client

“Understand difference between changing the past vs the future. If you have majority hashrate, you can decide future - not past.” -@aantonop

via Twitter Web Client

“Blockchains are only inherently tamper-evident. In order to be tamper-proof they need proof of work.” -@aantonop https://t.co/z7N64ZV0sa

via Twitter Web Client