Ethereum is truly innovative; the first genuine stab at a post-Bitcoin blockchain.
It's also going to take many, many years to work the kinks out. The scripting language is so much more sophisticated that it'll be quite some time before the ins and outs of attack vectors will be reasonably understood.
I'll save comments on the mining for another time; I'm not totally up to date with their current plans. One thing any mining strategy needs to have a solution for is the popularity / security dichotomy: Bitcoin has provably much more compute than you securing the network. This makes you want to tag on to Bitcoin, with, say the side-chain proposal.
On the other hand, if you do that, pre-mining becomes harder, and you will be dealing with ASIC cartel economics. If you go the other way and build a 'non-ASIC-able' proof of work, you probably don't want to accidentally get on the ASIC bandwagon when you've promised your customers that they won't be on it.
> If you go the other way and build a 'non-ASIC-able' proof of work, you probably don't want to accidentally get on the ASIC bandwagon when you've promised your customers that they won't be on it.
You also run the risk that anybody with access to a botnet might be able to gain 51% of the network's processing power.
You didn't actually answer any of my questions. Why is ethereum useful? It's also not that innovative. Increase the scope of the scripting language and charge for ops, and you've just turned Bitcoin into it.
You'd also need a key-value store linked to contracts, as well as the ability for contracts to call each other. Those are more difficult (though your main point stands)
I actually think it's unfortunate that a "pure" currency was the first application of the blockchain. Bitcoin's problems are around mindshare and social acceptance, because it's useless without it. Something like Filecoin interests me much more, because the value of the system is a bit more independent of crowd psychology.
It's also going to take many, many years to work the kinks out. The scripting language is so much more sophisticated that it'll be quite some time before the ins and outs of attack vectors will be reasonably understood.
I'll save comments on the mining for another time; I'm not totally up to date with their current plans. One thing any mining strategy needs to have a solution for is the popularity / security dichotomy: Bitcoin has provably much more compute than you securing the network. This makes you want to tag on to Bitcoin, with, say the side-chain proposal.
On the other hand, if you do that, pre-mining becomes harder, and you will be dealing with ASIC cartel economics. If you go the other way and build a 'non-ASIC-able' proof of work, you probably don't want to accidentally get on the ASIC bandwagon when you've promised your customers that they won't be on it.