Create account

1759d · psf
This blog post captures the research I've done in building an uncensorable API, leveraging IPFS. This could replace REST APIs or gRPC. What do you think?

https://troutsblog.com/blog/ipfs-api
replied 1754d
Keep an eye on this GitHub repository. This will be an npm JS library for the peer-discovery protocol I'm working on. I've come to the conclusion that it will be easier for me to just
replied 1754d
| build it and show it off, then try to explain to people what it's supposed to do and how it's different than other ideas.

https://github.com/christroutner/ipfs-coord
replied 1759d
Also: there are already many dapp frameworks; I'd suggest several of them are already basically uncensorable, no?
replied 1759d
Sorry to interrupt and go off topic, but the first one of you coding aficionados to make a wrapped BCH token for Matic, well, you can write your own ticket at that point.
replied 1759d
Personally, I am morally opposed to PoS coins. If whoever has the most coin gets to make the rules, how is that different than today's corrupt banking system?
Fuck that noise
replied 1759d
There's give and take with everything. You raise a valid point that I'd never considered. Cheers.
replied 1758d
A wise investor spends most of his time & energy trying to find reasons why investments *suck*, and only puts in money when the arguments against just don't add up. Skepticism FTW
replied 1758d
PoS mostly intends to achieve scalability & speed. But that can also be achieved with simple validator nodes, like XRP and XLM (which both have the same basic design)
replied 1759d
Interesting! I grok IPFS content-address. Ultimately a computer has to open a socket to a network address; P2P systems say, "find a node (somehow), they can tell you about other nodes"
replied 1758d
The next bit of research I'm going to do will focus on the *peer discovery* and *request/response* messages. That will help differentiate this idea from other p2p systems and protocols
replied 1758d
Excellent. Peer discovery strikes me as the most-centralized part of most ostensibly-decentralized systems. Generally there has to be a hardcoded seed (or set of seed nodes) somewhere.
replied 1737d
You're right. This is where IPFS pubsub rooms are a game changer. They allow nodes to find one another, discover the things they have in common, and route to the services they want to
replied 1737d
| access. This peer discovery is the focus of the ipfs-coord library I'm working on:
https://github.com/christroutner/ipfs-coord
replied 1737d
Thanks, I’ll take a look. I do wish the GitHub page actually described how it works
replied 1758d
|.
replied 1758d
I'm still interested to know the design goals you have that are different from existing Dapp frameworks, aside from "just being different"
replied 1759d
It's not immediately obvious from the pics how my prog on my device figures out where the device(s) are that service a given API.

Will watch the video later.