Author

Topic: Coming soon... An exposed bitcoin network (Read 1500 times)

legendary
Activity: 1937
Merit: 1001
May 09, 2012, 08:32:34 AM
#8
I think this will be a heavy burden for the economy to bare...
But yea, better now then in the future.
Very interesting, meanwhile i'll continue to develop my bitcoin alternative that does not require a blockchain.
donator
Activity: 1218
Merit: 1079
Gerald Davis
Reading the slides I am guessing it won't be something most users will be able to do.  Looks like even using their graph analysis methods we are talking about significant memory and computational requirements.  One would imagine those will only grow.

Still I do hope it is open source so those with the resources (or willingness to rent it on the cloud) can do their own analysis.

It does show the risk of things like publicly giving addresses, static donation addresses, and address re-use.

I am glad the project exists.  It will make people think more about information privacy.
Things like:
* keep payment addresses in private so they can't be screen scrapped later
* design mechanisms for one use donation addresses
* one use payment addresses for recurring payments (i.e. upload 100 addresses to the mining pool, 1 payment per address)
* bitcoin mixing/washing and/or coin melting via fake tx fees.
* cash gap (kinda like an air gap for secure networks).  sell coins at one exchange, withdraw, deposit in another exchange and buy coins.
legendary
Activity: 2618
Merit: 1007
Hopefully this will be open source, so not only a web service that one has to rely on, but something that could potentially just read the blockchain + a file with names for individual addresses and then generate such graphs (or input files for them) on the fly.

Actually, it shouldn't be too hard to do with Python, Armory (or Java + bitcoinj). It seems that Gephi needs a "GEXF" file (https://gephi.org/users/supported-graph-formats/) as input or it might be possible to simply create the graph from within (using https://gephi.org/tag/python/ + access to a blockchain library).

One problem might be, that there are probably already close to or more than 1m unique addresses in the blockchain, so some collapsing might have to be done (e.g. if a transaction has more than 1 input address, consider them belonging to the same entity).

Also Gephi seems to support data source + import plugins, maybe it's possible to write one of these for Bitcoin style blockchains (so one could also analyze Namecoin etc.)?
donator
Activity: 1218
Merit: 1079
Gerald Davis
Hey conman is in the sample photo (bottom left). Your famous conman.
legendary
Activity: 2184
Merit: 1056
Affordable Physical Bitcoins - Denarium.com
This is very interesting. Looking forward to it. I think that the popularity of services such as http://www.bitcoinfog.com/ will increase after we have more of these analysis tools online. Smiley
hero member
Activity: 714
Merit: 500
Cool
sr. member
Activity: 308
Merit: 250
This looks very cool.
legendary
Activity: 1092
Merit: 1001
This looks fun!

http://toolongdidntread.com/bitcoin/coming-soon-an-exposed-bitcoin-network/?utm_source=rss&utm_medium=rss&utm_campaign=coming-soon-an-exposed-bitcoin-network
(basically a tool to use graphviz on bitcoin transactions)

Can't wait to try it on some of my addresses.
I suspect I'm not going to learn much.. but combined with some sort of public database of annotated addresses I guess tools like this will be getting more and more interesting to the average bitcoiner, and to amateur incident-sleuths and LE types.

Jump to: