Author

Topic: Encrypted distributed file system. (Read 148 times)

member
Activity: 210
Merit: 26
High fees = low BTC price
March 06, 2018, 02:59:47 PM
#3
What are you trying to accomplish? Whats the reason?

Like taking back OUR internet from the government where everything has become hate speech
and anything that can be taxes is taxed.

Quote
You can't 'break down SHA-Hash into 2mb blocks'.
SHA is a hash function. Depending on which you choose it will create an output with the length X (e.g. 256 bit for SHA-256).
SHA is not used for encryption. Its a one-way function. Easy to calculate forwards but almost impossible backwards


Here I am talking about SHA-Hash being a hash of all the 2mb blocks making up the file much like used by Bit-Torrent
but maybe I didn't make it clear

Quote
An email? Roll Eyes
Yes total agreement, 100% right and this is why I use public keys throughout the network so I could
send you a WhatsApp type message without the source code being private and needing central servers.
It's a fair bet WA would had been closed down without coming to an agreement with the powers that be and
I trust them as far as I can spit myself.

Quote
For me this seems like merging a few altcoin projects to create one inconsistent theoretical project.   Undecided

Nothing wrong with being "theoretical" so that's why I am asking questions and it started off when I looked at crypto coins and realized like
many before me that we needed more of a money network than this coin or that coins and then I kind of went sideways and ended up with a
socket based network that instead of using IP/Port it uses Public-Key/Service so encrypted apps could be built on top of it.

The project really has nothing to do with alt-coins apart from fake "gas" and really does not improve much on Bit-Torrents
when it comes to movies but it does offer secure file transfer without big brother poking his nose in by using the same security that
you trust with your coins which is better than the encryption that was later bolted on to Bit-Torrent when ISPs started "Traffic Shaping"
my downloads.
 






legendary
Activity: 1624
Merit: 2481
March 06, 2018, 01:36:10 PM
#2
I have been working on node discovery that will scale to a hundred million users or more and it uses
public/private keys on a private network that are then used to sub-domain the network into areas of
responsibility so if the public key of a node is abc12345..... then this node would be a good starting point if you
are trying to find the ip-address for a node with the public key of ab99999........

What are you trying to accomplish? Whats the reason?



Now I am moving on to a files system that needs to be anonymous and encrypted so that anyone can publish a file
to the network and nodes that host parts of the file (Chunks in Bit-torrent terms) will not even know what they
are hosting because the data was encrypted before it was published.

My best plan so far is to SHA-Hash the whole file and then break it down into something like 2mb blocks and spread them
across the network using a system of PUSH,GOT,GET but again sub-domain responsibility based on the hash of the file plus
caching of file chunks that are popular and to use a system of relays nodes so that a no one knows if the request is direct (P2P) or
is via a relay node (Bit Like Tor)

You can't 'break down SHA-Hash into 2mb blocks'.
SHA is a hash function. Depending on which you choose it will create an output with the length X (e.g. 256 bit for SHA-256).
SHA is not used for encryption. Its a one-way function. Easy to calculate forwards but almost impossible backwards




Ultimately the data is useless without the decryption key that could be sent via an email privately (Secure details for blowing up the world)
or published via a type of torrent-site service that some nodes might be running.

An email? Roll Eyes



What I am trying to avoid is having to keep encrypting/decryption the the file chunks as they pass between relay nodes or are loaded from
file as this is CPU Intensive and involves too many handshakes on the secure network plus I am thinking of ...

The file should be encrypted/decrypted once. If every node would be able to decrypt a file.. what sense would the encryption then have?



so what do you think and can it be improved ?

For me this seems like merging a few altcoin projects to create one inconsistent theoretical project.   Undecided

member
Activity: 210
Merit: 26
High fees = low BTC price
March 06, 2018, 12:04:31 PM
#1
I have been working on node discovery that will scale to a hundred million users or more and it uses
public/private keys on a private network that are then used to sub-domain the network into areas of
responsibility so if the public key of a node is abc12345..... then this node would be a good starting point if you
are trying to find the ip-address for a node with the public key of ab99999........

Now I am moving on to a files system that needs to be anonymous and encrypted so that anyone can publish a file
to the network and nodes that host parts of the file (Chunks in Bit-torrent terms) will not even know what they
are hosting because the data was encrypted before it was published.

My best plan so far is to SHA-Hash the whole file and then break it down into something like 2mb blocks and spread them
across the network using a system of PUSH,GOT,GET but again sub-domain responsibility based on the hash of the file plus
caching of file chunks that are popular and to use a system of relays nodes so that a no one knows if the request is direct (P2P) or
is via a relay node (Bit Like Tor)

Ultimately the data is useless without the decryption key that could be sent via an email privately (Secure details for blowing up the world)
or published via a type of torrent-site service that some nodes might be running.

What I am trying to avoid is having to keep encrypting/decryption the the file chunks as they pass between relay nodes or are loaded from
file as this is CPU Intensive and involves too many handshakes on the secure network plus I am thinking of using artificial "Gas" so that
nodes gain credits for services provided to the network and will be placed at the back of the queue if they just act like parasites.

Time stamps will be used to age old files and to free up node disk-drive space so what do you think and can it be improved ?

   



 
Jump to: