Dnia niedziela, 6 lipca 2014 13:42:55 grarpamp pisze: > On Sat, Jul 5, 2014 at 5:35 PM, Lodewijk andré de la porte <[email protected]> wrote: > > Just write the communications layer. Expose it through a socket on the > > local machine. Let's say the port will be 33742 (actually a good port > > number) (say > > > > This little daemon/server maintains the connection to whatever P2P network > > is currently thought to be neat. It may or may not also do DNS-ish stuff > > and (if that's the popular P2P thing) it can have a keyring with trusted > > peers (aka "friends"). > > > > Once you have that you can communicate in an abstracted form. You can send > > Yes, I've always felt the level of interop among crypto darknets is > arbitrarily far less than ideal. Everybody seems to want to make their > own user protocol for umm, say, the simple act of sending a message. > And with all these different vertical darknet silos, you'll never be able to > seamlessly cross message your friends on some other nets you also happen to > be running. Then even if you got a pluggable *user* protocol for messaging, > surfing, storage ironed out and made, you still have the problem of > *backend* dest (and src) addressing. ie: There are at least four, maybe six > nets I know of that can present an IPv6 tunnel interface to the user. > Making it dead simple to securely route all your traffic into the > darknet[s], or the entire set of darknets > based on CIDR block addressing into them. But no, right now they overlap and > conflict :( And without IPv6, they use different native addressing schemes. > > This begs for a serious sitdown conference amongst darknets to see what > better cooperative user and backend interfaces might be possible. Or at > least create one grand DHT based middleware addressing shim between > them and the user (and preferably one that presents IPv6 tun interface, you > know... because every app on the planet can speak that these days. > Which is a big adoption win.)
Absolutely. Without serious and seamless interoperability, we have no way of actually getting people to use these solutions. "Nobody's there" × tens of solutions/networks. This is, by the way, a huge problem in the free and open decentralised/federated crowd. Consider this e-mail exchange: http://lists.w3.org/Archives/Public/public-fedsocweb/2013May/0058.html -- Pozdr rysiek
signature.asc
Description: This is a digitally signed message part.
