Hacker News Reader

Comment by Retric | Hacker News Reader

That assumes it survives with Ethereum's growing market share and better technology Bitcoin could simply die at some point in the next decade.

The problem is once the number of miners drops enough someone,likely with a strong stake in an Altcoin, can simply buy enough hashing power to just take over for a few weeks resulting in an abrupt end.

View on HN

Some of Ethereum's tech may be better, but their Windows client is lacking in terms of customization. If you don't have enough free space on your C drive, you can't run it... at least without symlinks or manually running geth with command line options first before running the ethereum client, which is very kludgey and doesn't work well with the internal updater. Bitcoin, Litecoin, Dogecoin, etc all let you select the location of your blockchain store so a typical PC configuration of a fast smaller SSD C: and larger D: etc data store can run it without issue. And you can install those clients anywhere you like and keep the data anywhere you like for maximum customization and security.

Their GUI Linux clients weren't much better when I checked a year ago. But this is the kind of thing that will probably work itself out, being open source and all.

I see comments like this echoed whenever bitcoin is mentioned outside of the usual circles. I always assume that the individual is personally invested in ethereum's success and/or they don't understand the technology.

Currently, ethereum has inherited all of the problems that bitcoin has and and magnified them by a substantial amount. We'll either witness the development team pull off a miracle - or more likely, in my opinion, watch the house of cards come crashing down.

Ethereum has the same issue practically all current blockchains have. In order to scale, payment channels need to be used. This is why Ethereum has run into backlogs lately and has its own version of Lightning on its roadmap. Bitcoin already has usable payment channel implementations ready and waiting for the Segregated Witness bug fix to activate.

Having ~60 times as many blocks / hour means you can potentially secure transactions vastly faster using Ethereum’s even if the total number of transactions does not change. So, even if nothing else changes this really is a significant advantage.

Now, they are hitting scaling issues. But it will be interesting to see if they are more responsive.

The GHOST protocol only made the security properties of a group of short block times similar to longer block times. Before GHOST, short block times had worse security properties. POW over time still determines how secure a transaction really is.

https://eprint.iacr.org/2013/881.pdf

https://github.com/ethereum/wiki/wiki/Design-Rationale

https://blog.ethereum.org/2015/09/14/on-slow-and-fast-block-...

The attack model changes based on transaction size. For sub 10,000$ transactions the threat model is not someone with more hashing power than the network. It's someone with 1/k th the networks processing power getting lucky. Which is bound by both total network hashing power and the number of blocks you wait giving a significant usability advantage to faster blocks.

That assumes the person actually needs fast block confirmation times. The security properties of block times gets pretty complex and bitcoin's model is no slouch: https://eprint.iacr.org/2015/1019.pdf

Reply on news.ycombinator.com