Some interesting points about this topic are made by @LynAldenContact in lynalden.com/proof-of-stake/. Going to summarize a few of my takeaways RE the question of blockchains as political systems 1/
One of the points Lyn makes and which I got in the replies is that blockchains are more of a market-choice system
1) users choose which chains to use, 2) users choose which forks to use (which is kind of the same thing as which chain), 3) users choose which software to use
2/
Point 2 is especially relevant because any breaking change to a protocol is a hard fork. If end-users don't go along with the fork, then you essentially have no power. 3/
This is something Ive been stewing on for years; heres my latest attempt and Id like to hear thoughts
A blockchain is a political system for operating a shared dataset which is constitutional (operation constrained by rules) and legislative (operation changes by rule change). 1/
Decentralized consensus is generally the mechanism for upholding the active ruleset. Because it's difficult to assert a new state that differs from the majority of other miners' expectations, the miners operating the network are bound to the rules. Thus: "constitutional." 2/
You can, however, change the operation of the network by changing the rules. Thus: "legislative."
The system for deploying changes to the rules determine the legislative process and resulting political system. 3/
P2P, distributed data networks like Hypercore and IPFS are basically BitTorrent variants
The whole idea is you can share files, databases, and logs using 1) a cryptographic URL, and 2) bandwidth-sharing, where downloaders can become seeders
In contrast to HTTP, Hyper and IPFS don't associate data with a particular service or dot-com. The URLs use public keys or content hashes instead of IP addresses.
It's also possible for anybody to rehost ("seed") the data.
This gets into the weeds a bit, but for anybody following the ecosystem, here are some bits that you might find interesting:
In the past, Hypercore has used a daemon called "hyperspace." This was used to centralize all the data and networking logic, which apps would then share.
Well: It looks like that won't be necessary anymore! Apps can just run Hypercore and its network themselves.
That's possible because the networking layer has gotten a lot "cheaper" to use.
The upside: no cumbersome daemon management, and local apps share data the same way that remote apps share data. It's all down to "do you have the keys" and "are you in the multiwriter pool"
How could Clubhouse turn around its growth trajectory?
I'm not much for analysis threads but I've been following this closely and I'm curious about how new networks can (or can't) challenge the incumbents.
I'm going to push through feeling self conscious about doing this thread and just share. I hope you find it interesting!
Contrary to popular belief, I don't actually think CH's biggest risk is being copied by existing social networks. I'm honestly starting to think Stories was an outlier. Just look at attempts to copy TikTok; there's obviously more to it than gluing a UX to existing graphs.