If you take a step back and look it from the context that the data is somewhat being served by untrusted distributed devices in a verifiable way; it’s pretty fr

If you take a step back and look it from the context that the data is somewhat being served by untrusted distributed devices in a verifiable way; it’s pretty freaking amazing it works at all! Sure a raspberry pie could replace this in heartbeat but it’s not about current throughput being good, it’s about making it work as intended, ensuring the design is theoretically scalable, and marching toward a fully distributed network. I realize the author is talking mostly about performance but the think that is exciting about ipfs is it’s not some monolithic project trying to take over the world but rather a modular set of standards (IPLD, MULTICODE, libp2p) that generally do a very good job of integrating with legacy and have reasonable future proofed design. Ipfs is just one application bring all these thing together is in a particular confirmation but many more interoperable variations are likely soon to come.