4 ms·
If you start a new blockchain then you probably don't need as hard proof-of-work problems as there is now in bitcoin. One thing that the article is missing to
by chvid 9y ago
If you start a new blockchain then you probably don't need as hard proof-of-work problems as there is now in bitcoin.
One thing that the article is missing to discuss is that for a blockchain to truly work there needs to be a reward system for the nodes to participate in the network (basically to cover the cost of this crazy expensive infrastructure). Ie. a cryptocurrency that can be mined.
- fiatjaf 9y agoThis. I always thought that any blockchain could only work if there was a reward system (otherwise you're dealing with trusted parties and thus no blockchain in the strict sense is needed), and thus it never made any sense to me to see people implementing blockchains everywhere. News: "Company X is using blockchains internally." Me: "Does that mean Company X have created a token that can be mined by its own computers, in a crazy internal competition that will waste thousands in energy consumption and hardware?"
- lagadu 9y agoCryptocurrencies don't necessarily have to be mined. Some newer ones have systems that aren't computationally intensive; they do that by effectively inverting the system of responsibility: they work by saying "if you hold our token, you have a vested interest in the integrity of the chain, therefore we trust you to validate transactions for the network" where "validate" means "vote on whether a transaction is legitimate". It's fairly interesting and still under research but on some smaller cryptocurrencies it seems to be working. Google "proof of stake" for more. edit: for a single organization that might not be very useful, as you already have one big central entity that you trust (the organization itself) but it is potentially an additional layer of security that can be tailored to the organization's needs.