3 ms·

(I'm explaining the dream here. If I knew what I was talking about, I would have done it.) If the blockchain community gets the contract/language issues sorted

(I'm explaining the dream here. If I knew what I was talking about, I would have done it.) If the blockchain community gets the contract/language issues sorted out, a small dev team could knock out a first-stage system like the fisheries system in three months and in two layers of technology (contracts, oracles). It would be trivial to operate and resilient to server failure (less-so the oracles). Ten years later, it would still be just the same two layers of code. In this world, contracts make custom APIs obsolete. There is a new career path for a developer who lives and breathes async contract code. The way of implementing such a system now is db-centric. Business problems tend to be event-driven, but databases are not. So we need many more layers of technology: ERD, stored procedures, 'backend', partner API, hosting complexity, business continuity complexity, further layers to assist support and deployment and API onboarding. The database company takes on a life of its own, it is expensive to fund and delivers a bad customer experience. So the thing that blockchain adds: you will able to reliably build significant systems with two-person teams in domains where we currently struggle to do adequate work with firms of twenty or forty people. We will be able to engage with more complex domains than we can at the moment, and there will be network effects from this.

Jasper_ 9y ago
So it sounds like what you really want is a better centralized database that has those properties, rather than a slow, energy-inefficient ledger providing features you don't need like reward tokens, mining, and decentralization.
- cturner 9y ago
  You don't have mining in proof-of-stake. (Mining is for proof-of-work.) My focus is multiple-host, centralised async systems. I think this is the future, and that databases will be marginalised. Comment here elsewhere gives more details. I may turn out to be wrong. In that case, we will end up in the kind of blockchain world described in my earlier comments. I suspect it will hinge on whether it is possible to create a good-enough high-level language.

Jasper_ 9y ago

So it sounds like what you really want is a better centralized database that has those properties, rather than a slow, energy-inefficient ledger providing features you don't need like reward tokens, mining, and decentralization.

cturner 9y ago
You don't have mining in proof-of-stake. (Mining is for proof-of-work.) My focus is multiple-host, centralised async systems. I think this is the future, and that databases will be marginalised. Comment here elsewhere gives more details. I may turn out to be wrong. In that case, we will end up in the kind of blockchain world described in my earlier comments. I suspect it will hinge on whether it is possible to create a good-enough high-level language.