loading...

On Decentralization and "Next Gen Web" Wannabes

louy2 profile image Yufan Lou ・2 min read

Decentralization is not a virtue per se. The Web did not become ubiquitous and successful because it was decentralized. Another attempt at decentralization for decentralization's sake will not succeed.

In the context of early internet (ARPANet), decentralization meant eliminating single points of failure, in the face of a nuclear threat. In the context of bitcoin, decentralization means eliminating single points of responsibility, in the face of legal threats. In the context of smart contracts, decentralization means eliminating enforcement bottlenecks, in the face of shortage of court resources. In each case decentralization serves a use case other than for the sake of decentralization.

What has made the world wide web as we know it so successful? It was not the fact that it could still remain in operation after a nuclear strike. Ironically now we have built such a brittle network that a BGP configuration mistake can cripple half of it. No, it was Netscape and Internet Explorer, popularizing web browsing. It was AOL, Yahoo, Google, Wikipedia, helping laypeople reach into and make use of the information on the WWW. It was Adobe Flash, Chrome, iPhone, pushing the limit of what use cases the WWW can deliver.

IPv6 had all the resources in the world to replace IPv4. To this day IPv6 adoption has just climbed to around 40%, mostly in internal networks, and IPv4 has seen no decline, with prices for IPv4 addresses increasing. That was an official effort to replace the Internet. Nothing else has even come close to it. Ethereum is still an experiment after 4 years. IPFS is a glorified archive.org. Is anyone still using Namecoin? How many know ZeroNet?

Research it, push the distributed system frontier forward, cool. Like ZeroTier, maybe you can make a better VPN. Just don't expect it to be "the next generation internet".


Originally under:

at:

Decentralization is not a virtue per se. The Web did not become ubiquitous and successful because it was decentralized. Another attempt at decentralization for decentralization's sake will not succeed.

In the context of early internet (ARPANet), decentralization meant eliminating single points of failure, in the face of a nuclear threat. In the context of bitcoin, decentralization means eliminating single points of responsibility, in the face of legal threats. In the context of smart contracts, decentralization means eliminating enforcement bottlenecks, in the face of shortage of court resources. In each case decentralization serves a use case other than for the sake of decentralization.

What has made the world wide web as we know it so successful? It was not the fact that it could still remain in operation after a nuclear strike. Ironically now we have built such a brittle network that a BGP configuration mistake can cripple half of it. No, it was Netscape and Internet Explorer, popularizing web browsing. It was AOL, Yahoo, Google, Wikipedia, helping laypeople reach into and make use of the information on the WWW. It was Adobe Flash, Chrome, iPhone, pushing the limit of what use cases the WWW can deliver.

IPv6 had all the resources in the world to replace IPv4. To this day IPv6 adoption has just climbed to around 40%, mostly in internal networks, and IPv4 has seen no decline, with prices for IPv4 addresses increasing. That was an official effort to replace the Internet. Nothing else has even come close to it. Ethereum is still an experiment after 4 years. IPFS is a glorified archive.org. Is anyone still using Namecoin? How many know ZeroNet?

Research it, push the distributed system frontier forward, cool. Like ZeroTier, maybe you can make a better VPN. Just don't expect it to be "the next generation internet".

Posted on by:

louy2 profile

Yufan Lou

@louy2

Learning Rust and Haskell, tired of JavaScript. Thinks Ruby is awesome except that it is not cross-platform enough. 日本語 / 中文 OK. He/him.

Discussion

markdown guide