@roboneko @cy@mstdn.io @Hyolobrika@mstdn.io The blockchain solves Zooko's exactly the same way NFTs solve ownership.
@Hyolobrika @roboneko The thing is, domain names are merely a convention we take for granted. And service providers take for granted that we do.
We could have been putting PKEYs in URLs. Well, that's what TOR "hidden" sites do anyway, without any ICANN, ICAN'T or IFORCEU whatsoever.
Also, consider magnet links, ipfs links, git hashes, etc etc
@roboneko @cy@mstdn.io @Hyolobrika@mstdn.io Anonymity is on demand in the services. You can specify amount of hops or traffic to mix with. You don't need distinct TOR and plain TCP protocols, it's just a connection option.
@roboneko @cy@mstdn.io @Hyolobrika@mstdn.io Discovery is pretty long right now, tbh. They need way more participating nodes. But with a CADET circuit established, onion or not, you can have it interactive. I doubt it would work for twitch shooters, but something more reasonable can be fine.
Streaming DHT propagation can be quick too, after some warm-up.
@xenmen Too bad computers making decisions will eat competitors for lunch. (And then, the rest of the world.)
@cy@mstdn.io @roboneko @Hyolobrika@mstdn.io DNS is just a common protocol for apps to resolve keys into values, with features like delegation and TTL. Same with GNS. You can have a bunch of TOR addresses pinned by your own private key, accessible from anywhere on the network. Or files. Or app-specific data.
ReclaimID is the inverse of "Login with Facebook". It is "you can ask me if I still want to share stuff with you".
So, those aren't just some configuration pieces exposed to as JSON. But you'll have to read up some nice materials to make sense of the data.
@cy@mstdn.io @roboneko @Hyolobrika@mstdn.io huh, what?..
@cy@mstdn.io @roboneko @Hyolobrika@mstdn.io DoH is DNS-over-HTTP, a tech to stop providers and pretty much everyone else to snoop on you and poison your query results.
re:claimID is like reverese OpenID. You publish the attributes and the servers are requesting them to authenticate you.
@Hyolobrika@mstdn.io The web
links are name-centric, pointing to whatever the content is currently provided by whatever the authority managed to receive your request. Ephemeral and basically worthless. "Good URLs don't break", hear, hear. PKI is FUBAR, servers can 4xx and 5xx at you at any time or go black entirely. It's a little miracle we can serve stuff on localhost, and a divine intervention that the big web works at all, at least sometimes.
@mitchconner @Hyolobrika@mstdn.io @Moon
> .BIT DOMAINS
> Decentralized domains using
Namecoin cryptocurrency.
No.
@Hyolobrika@mstdn.io There are no TLDs in GNS. Not like ICANN runs it. Every PKEY is its own TLD.
Tell me your PKEY, and I put it under whatever the name *I* want. You don't even have a say in that. Think user bookmarks, not global registry of stuff.
If you have to tell me about some stuff you publish, you suffix it with your PKEY. But most of the time, GNS isn't a right level of abstraction to work with.
@cy@mstdn.io @Hyolobrika@mstdn.io @roboneko You don't need a REST API to fetch files. If anything, REST, HTTP even is unsuitable for high-latency P2P fetching.
If you want a browser running on a such a network you can invoke `gnunet-download` to crawl the URLs you encounter.
Toots as he pleases.