2
1
I'm looking for information on how Namecoin could create a "Facebook-style" social network and what prior work exists.
I understand that namecoin has fundamental problems and fledgling support, but I'm interested in where this idea has gone in terms of creating a distributed social network.
1
Well Retroshare somewhat falls into this category. It is not Coin-Based however, rather related to Bittorrent DHT technology und not well documented.
There is also the Bitmessage protocol which uses Coin technology for exchanging mail.
I do not know why you would build such a system as a Coin network, plain BT style DHT is more suitable for this purpose in my opinion. I do not know any good reason why Bitmessage does it this way either. I think it is just overcomplicated and not well thought through. The usecase is just very different from Bitcoin.
I mean, what would be the purpose of the block chain? Prevent double friending?
0
I wrote a bitmessage based, distributed, facebook a few months back. It used namecoin as the database to store everyone's statuses and comments and such. Contact me at this bitmessage if you are interested in it: BM-2cXji1uDyuXT94AoUG2u3GCApSFdbtri33
How would a DHT system without mining prevent nodes from hosting dishonest or incorrect data? – bvpx – 2013-10-21T21:00:27.193
By cryptographically signing this data as with pgp email. Retroshare implemets this (with pgp even). Bitcoin transactions are also signed btw. The block chain serves an additional purpose, namely to prevent double spending by cryptographically sealing a transaction log, so that the same coin cannot be signed and sent twice. – Paul Hänsch – 2013-10-21T21:03:18.993
Couldn't anyone then post the same message you posted again, and no one would be able to tell if you wanted to post again or if someone else posted again for you? "Post" refers to any action taken that's recorded on the network. – bvpx – 2013-10-24T21:38:43.963
You would simply include a timestamp into the signed message. This is standard procedure in Kerberos and other cryptographic systems. A replay of the original message would have to retain the original timestamp to match with the signature. Even a simple time independent message ID would be sufficient to sort out duplicates. Again this is a prooven procedure and block chaining is just ridiculously overcomplicated for this. Again the requirement in Bitcoin is different, because in Bitcoin, the original poster could be the "attacker". – Paul Hänsch – 2013-10-24T21:45:59.890
Could you explain why adding a timestamp to the signed message isn't enough to prevent double spending in Bitcoin? – bvpx – 2013-10-24T21:53:18.987
Because it does not prevent the original payer to replay a transaction with a new time stamp. After all the sender does have the keys for signing the message. Please read the original bitcoin paper on http://bitcoin.org/bitcoin.pdf or open another question, as this goes beyond the topic.
– Paul Hänsch – 2013-10-24T21:59:38.907