On Fri, Jan 19, 2018 at 02:11:37PM +0100, marc wrote: > Quoting and compelementing kenneth points below. But first, > top-posting since i feel it a new thread :!
Er wait.. you are calling the thread "distributed" while talking about blockchain broadcasts? I wouldn't call inefficient network-wide consensus operations "distributed". They are broadcasts plus agreement overhead on top. A proper architecture delivers events only to those who subscribed to have them, ideally using distribution trees. Why should anyone want to send every tweet to every possible Twitter user? Even Twitter uses optimized distribution trees! > By the way, how to make any action to rely in more local consensus, > instead of all those global ones? in this regard there is this 4 > minutes interview with Mark S. Miller > https://www.youtube.com/watch?v=21R259Wei2w I didn't hear him speak of local consensus. Whatever that is, it doesn't sound like it is limited to exactly the people who subscribed certain content, so it will probably still consume too much electricity, planet and bandwidth. Bitmessage uses the "stream" strategy to reduce overall traffic. It is still far too much, but at least it introduces pretty good metadata protection. So you may want to look into Bitmessage for social networking. > >you’ll store your profile/user data (social graph) on blockchain, not > >a service (like uport), > > this is the basic dependence, for starting scaling up, so up to what > points needs to or can be extended or needs which additional > hardcoding? How can anything that goes onto the blockchain scale up? > >You’ll have !groups on a colony/district0x/aragon chain, > > forums and mailing lists as well? ;) Reinventing Bitmessage? > >ipfs/swarm/storj will hold the content, > > what about steem.io? Which introduces ulterior inefficiency. Instead of pushing just the content to the subscribers you are telling them to go fetch it from some other overhead- prone technology. > >basic income might come from Proof of Presence, being online, being a > >node, creating the network, > > i guess you mean "content curator". however, if he is giving his > interactive data to the commons-system, an evolved > ArtificialGeneralInteligence could use the browsing timings as > decision (if you stay more than 5 senconds watching something it > gets x value.., etc) You would want all users to publish how long they have been watching some crp and prove it by some cryptographic means? Just when I thought we had enough surveillance economy. > otherwise please share more info about such proof-of-presence concept. > > presence is a very mergy concept, not enough used in neuroscience > nowadays.. Has anyone told you blockchain people that basic income models over ethereum fail because the super-rich needed to finance it will simply not take part? Maybe it takes a certain lack of rationality to successfully become a fan of blockchains. > >I guess the activity stream also works decentralized: > >instead of the list of posts/activity living on server, every single > >post would be chained to a > >previous post as a meta-data crypto-pair. So looking at your feed would > >look like etherscan, > >that lists the contract's entire transaction history. Who pays for the electricity bill, considering that most activity streams are utterly boring for the majority of mankind? Is it really so disturbing to use a protocol that was actually designed to do these things instead? Eight years ago there was this "gnu social p2p" vs "gnu social" dispute. Teddks was the man who convinced me that federation was a dead horse. Now we are dealing with dead elephants as an alternative to it. Progress! -- E-mail is public! Talk to me in private using encryption: http://loupsycedyglgamf.onion/LynX/ irc://loupsycedyglgamf.onion:67/lynX https://psyced.org:34443/LynX/
