I spent about three months consulting for some crypto company as a distributed systems engineer.
They paid me to design an eventually consistent, self-healing data store with a cache layer / write ahead log, with peers determined by paxos consensus, transfers metered by finops and govered with kademlia, and a storage layer capable of byzantine fault tolerance, which we implemented via signature chains.
See, they had this crazy idea that they'd make a cryptocurrency that they could sell to western digital, who could offer hard drives that "filled themselves up" with other people's data when idle. WD would obv make a buck and maybe sell these drives for much cheaper than the component cost. I'm not exactly sure of the economics. I think the idea was to have half the drives be "receivers" and half be "senders" and actually sell the "senders" for way more than component cost, but provide trivial effort file backup.
I think they're still building it. I dunno if it'll be a scam or not. I had fun though.
They paid me to design an eventually consistent, self-healing data store with a cache layer / write ahead log, with peers determined by paxos consensus, transfers metered by finops and govered with kademlia, and a storage layer capable of byzantine fault tolerance, which we implemented via signature chains.
See, they had this crazy idea that they'd make a cryptocurrency that they could sell to western digital, who could offer hard drives that "filled themselves up" with other people's data when idle. WD would obv make a buck and maybe sell these drives for much cheaper than the component cost. I'm not exactly sure of the economics. I think the idea was to have half the drives be "receivers" and half be "senders" and actually sell the "senders" for way more than component cost, but provide trivial effort file backup.
I think they're still building it. I dunno if it'll be a scam or not. I had fun though.