Bitcoin Forum
December 06, 2016, 04:24:55 PM *
News: Latest stable version of Bitcoin Core: 0.13.1  [Torrent].
 
   Home   Help Search Donate Login Register  
Pages: [1]
  Print  
Author Topic: Idea: Bitcointorrent  (Read 1911 times)
jpent
Jr. Member
*
Offline Offline

Activity: 31



View Profile
April 24, 2011, 04:42:04 PM
 #1

Hey.

This is an idea I had while in a discussion on #bitcoin-otc. I mentioned it there not entirely seriously but some people seemed to think it could be worth something.

So the idea is this:

With bittorrent, as I understand it, all the files are divided into 'chunks' which represent a small portion of data in a file. These chunks are sent from people who have them to people who don't, eventually meaning that everyone in the swarm should get the entire file. The problem is that there is not usually an incentive to seed chunks after you've got everything you want. You may also want to limit the speed at which you seed, although usually if you refuse to seed at all while downloading you will be boycotted.

Now imagine if for every chunk you receive from someone, you were required to send a fixed sum of BTC, say 0.001. This would mean for a download of 1000 chunks, you would end up paying the various seeders a total of 1 BTC. However, as you are seeding to others, you will also receive BTC in return. This would give you and others an incentive to seed as much as possible before and after the download, which could very will give a performance boost.

This might seem like a pyramid scheme, and to some extent it is. No matter what, some people will not be able/will not want to seed to 100% and end up paying something for the torrent. However, in return they should get a faster download.

There are some technological issues also. Each time a chunk is sent, a trade will have to occur, chunks for bitcoins. With any trade on the web it is possible that one party will not honour the trade. Also, it would require many successive micropayments, so it would have to be done without using normal bitcoin transactions, and without increasing bandwidth overhead too much.

Anyway I just thought I'd put the idea out there to see what people thought of it. I'll know in a month or so whether I'll have time to work on it myself, so if by that time I do, and still think it's worth it I'll probably start some more serious research. Also, apologies if I've completely misunderstood how bittorrent works...

PGP Key Fingerprint: 142E BEF5 5420 B00F 186C  7332 0B15 F673 EF9F DA26

# make install, not war

149gNEaA45b4NkM37XuBbhrM3gTAKikzqd - In case anyone needs this
1481041495
Hero Member
*
Offline Offline

Posts: 1481041495

View Profile Personal Message (Offline)

Ignore
1481041495
Reply with quote  #2

1481041495
Report to moderator
Advertised sites are not endorsed by the Bitcoin Forum. They may be unsafe, untrustworthy, or illegal in your jurisdiction. Advertise here.
SmokeTooMuch
Legendary
*
Offline Offline

Activity: 873


View Profile
April 25, 2011, 12:47:53 AM
 #2

Hmm,
what happens with unfinished seeds ?

Let's say someone is sharing a file with 10 chunks, but only 9 could be seeded to the network.
If I am trying to download this file I would credit the people for downloading the 9 chunks but could never get the 10th.
This way I'd be spending money on data garbage.

I like the idea, but I'm not sure if Bitcoin is the right thing for that.
There should be some "traffic credit" which doesn't have any other value than spending for traffic.


Actually I'd still like to see a revive of BitBlinder.
http://torrentfreak.com/free-anonymous-bittorrent-becomes-reality-with-bitblinder-090611/

Date Registered: 2009-12-10 | I'm using GPG, pm me for my public key. | Bitcoin on Reddit: https://www.reddit.com/r/btc
You like what I'm doing? Why don't you send me a coin: 17Pj8jpUgY6qTaKgiopL5U48zxU4rTrkuB
bitcoinBull
Legendary
*
Offline Offline

Activity: 826


rippleFanatic


View Profile
April 25, 2011, 02:21:02 AM
 #3

I posted before in a thread about "btc2" a similar idea.

The idea is that where bitcoin converts CPU/GPU power to coins, an alternative would be to convert network bandwidth to coins.

I'd prefer to implement this coining of bandwidth in Tor, so it could be called Torcoin.  And because there would be consistent activity in an effort to mint torcoins, this might help protect against some of the correlational vulnerabilities which can defeat tor anonymity.  But this is just a guess, I'm no tor expert.


College of Bucking Bulls Knowledge
FooDSt4mP
Full Member
***
Offline Offline

Activity: 182


View Profile
April 25, 2011, 08:51:00 AM
 #4

I posted before in a thread about "btc2" a similar idea.

The idea is that where bitcoin converts CPU/GPU power to coins, an alternative would be to convert network bandwidth to coins.

I'd prefer to implement this coining of bandwidth in Tor, so it could be called Torcoin.  And because there would be consistent activity in an effort to mint torcoins, this might help protect against some of the correlational vulnerabilities which can defeat tor anonymity.  But this is just a guess, I'm no tor expert.




I understand how to use a GPU to secure a blockchain... how do you do the same with bandwidth?

As we slide down the banister of life, this is just another splinter in our ass.
jpent
Jr. Member
*
Offline Offline

Activity: 31



View Profile
April 25, 2011, 12:01:32 PM
 #5

Let's say someone is sharing a file with 10 chunks, but only 9 could be seeded to the network.
If I am trying to download this file I would credit the people for downloading the 9 chunks but could never get the 10th.
This way I'd be spending money on data garbage.

Hmm. Maybe it would be possible for chunks to valued differently, creating an incentive for people to try to find the missing chunks and seed them? This does add a layer of complexity though.

I'll check out bitblinder later.

PGP Key Fingerprint: 142E BEF5 5420 B00F 186C  7332 0B15 F673 EF9F DA26

# make install, not war

149gNEaA45b4NkM37XuBbhrM3gTAKikzqd - In case anyone needs this
Insti
Sr. Member
****
Offline Offline

Activity: 294


Firstbits: 1duzy


View Profile
April 25, 2011, 12:28:04 PM
 #6

This has been discussed before. Search the archives.

The general conclusion I think was, that bitcoin is not really the system you should be using for doing lots of microtransactions. So you end up having to bundle your transactions up into larger transactions which means you have to do your own chunk/packet accounting anyway. So you end up with a system that is quite similar to existing ratio type systems. But which allow you to buy your way to a better ratio with bitcoins. (Which is still not a bad idea.)


ShadowOfHarbringer
Legendary
*
Offline Offline

Activity: 1470


Bringing Legendary Har® to you since 1952


View Profile
April 25, 2011, 01:41:51 PM
 #7

This has been discussed before. Search the archives.

The general conclusion I think was, that bitcoin is not really the system you should be using for doing lots of microtransactions.

+1

And also -1 to any unnecessary bloat which would put more strain on the network.

Shortline
Full Member
***
Offline Offline

Activity: 123


View Profile
April 26, 2011, 04:52:06 AM
 #8

If you could escape from the seemingly limitless reach of litigious media rights-holders, a tracker that pays people btc to seed files would be pretty kick-ass. Just make people pay for the last 10% of any given file to motivate people to seed to completion. The bounty goes up for every month nobody DLs a torrent or every seeder that drops out, goes down when it is downloaded to let leechers make their money back, and the website has a discount section for seeders who feel like they're going to give up on seeding a file.

But you'll never escape if you start actually facilitating payments between filesharers.

bitcoinBull
Legendary
*
Offline Offline

Activity: 826


rippleFanatic


View Profile
April 26, 2011, 06:59:12 AM
 #9

I posted before in a thread about "btc2" a similar idea.

The idea is that where bitcoin converts CPU/GPU power to coins, an alternative would be to convert network bandwidth to coins.

I'd prefer to implement this coining of bandwidth in Tor, so it could be called Torcoin.  And because there would be consistent activity in an effort to mint torcoins, this might help protect against some of the correlational vulnerabilities which can defeat tor anonymity.  But this is just a guess, I'm no tor expert.




I understand how to use a GPU to secure a blockchain... how do you do the same with bandwidth?


Well that's the question now isn't it.. I'm just hypothesizing that its possible, not making a concrete proposal.

So bitcoin mining expends CPU power with a probability of processing the correct block-chain hash for the next block in the chain.

A vague and non-technical proposal for torcoin:  

Torcoin mining expends bandwidth with the probability of transferring/processing the correct (not CPU intensive hash of) the next packet-block in the packet-chain.  So more bandwidth gives more probability of getting the key packet.  Tor exit nodes would have an advantage over non-exit nodes in transferring these "solution" packets, since more packets travel through exit nodes (at least when accessing external/non-hidden services).  This would give more incentive for the operation of exit nodes.  Packets from tor users would be hashed by the proxying nodes attempting to mine the torcoins in the next packet-chain.  The critical necessity for this system to function is that CPU power won't help find the next packet-block in the packet-chain, only bandwidth capacity.

Additionally, torcoins could be spent by tor users for increased priority for their packets.  So more torcoins spent by a user would enable greater tor network speed.  And inversely, the operator of a tor hidden service would mine more torcoins, thus giving incentive to operate popular tor hidden services.  Tor users without torcoins would get the default low speeds, but their packets would provide new torcoins for the network.  So maybe the total amount of torcoins in circulation would increase proportionally to the bandwidth transferred within the Tor network, inflationary style.  Or maybe the total torcoins in circulation could be limited absolutely, in the deflationary style of bitcoin.

The main goal here would be to monetize anonymous bandwidth, the same way bitcoin monetizes CPU power.




College of Bucking Bulls Knowledge
grue
Global Moderator
Legendary
*
Offline Offline

Activity: 1932



View Profile
April 26, 2011, 11:50:58 PM
 #10

I posted before in a thread about "btc2" a similar idea.

The idea is that where bitcoin converts CPU/GPU power to coins, an alternative would be to convert network bandwidth to coins.

I'd prefer to implement this coining of bandwidth in Tor, so it could be called Torcoin.  And because there would be consistent activity in an effort to mint torcoins, this might help protect against some of the correlational vulnerabilities which can defeat tor anonymity.  But this is just a guess, I'm no tor expert.




I understand how to use a GPU to secure a blockchain... how do you do the same with bandwidth?


Well that's the question now isn't it.. I'm just hypothesizing that its possible, not making a concrete proposal.

So bitcoin mining expends CPU power with a probability of processing the correct block-chain hash for the next block in the chain.

A vague and non-technical proposal for torcoin:  

Torcoin mining expends bandwidth with the probability of transferring/processing the correct (not CPU intensive hash of) the next packet-block in the packet-chain.  So more bandwidth gives more probability of getting the key packet.  Tor exit nodes would have an advantage over non-exit nodes in transferring these "solution" packets, since more packets travel through exit nodes (at least when accessing external/non-hidden services).  This would give more incentive for the operation of exit nodes.  Packets from tor users would be hashed by the proxying nodes attempting to mine the torcoins in the next packet-chain.  The critical necessity for this system to function is that CPU power won't help find the next packet-block in the packet-chain, only bandwidth capacity.

Additionally, torcoins could be spent by tor users for increased priority for their packets.  So more torcoins spent by a user would enable greater tor network speed.  And inversely, the operator of a tor hidden service would mine more torcoins, thus giving incentive to operate popular tor hidden services.  Tor users without torcoins would get the default low speeds, but their packets would provide new torcoins for the network.  So maybe the total amount of torcoins in circulation would increase proportionally to the bandwidth transferred within the Tor network, inflationary style.  Or maybe the total torcoins in circulation could be limited absolutely, in the deflationary style of bitcoin.

The main goal here would be to monetize anonymous bandwidth, the same way bitcoin monetizes CPU power.




what's preventing me from setting up 10 computers, with 10 GB/s connections (to each other) to "simulate" traffic?

It is pitch black. You are likely to be eaten by a grue.

Tired of annoying signature ads? Ad block for signatures
nphard
Member
**
Offline Offline

Activity: 67


View Profile
April 27, 2011, 09:42:51 AM
 #11

The general conclusion I think was, that bitcoin is not really the system you should be using for doing lots of microtransactions.

I really don't see the problem with this, and I really like this idea.  Sure, there would be a ton of  microtransactions, but they would be handled by the site.  Think of how betco.in works.  You wouldn't be flooding the bitcoin network, you would only actually withdraw bitcoins occasionally when you want to, instead of after every chunk of a torrent you're seeding.
Pages: [1]
  Print  
 
Jump to:  

Sponsored by , a Bitcoin-accepting VPN.
Powered by MySQL Powered by PHP Powered by SMF 1.1.19 | SMF © 2006-2009, Simple Machines Valid XHTML 1.0! Valid CSS!