4 ms·
We acquired the complete state of the Bitcoin transaction system [...] This required downloading 180,001 separate but linked HTML fi les [...] following the
by oakwhiz 13y ago
We acquired the complete state of the Bitcoin transaction system [...]
This required downloading 180,001 separate but linked HTML fi les [...]
following the links backwards to the zeroth block [...]
Each fi le was parsed in order to extract all the multisender/multireceiver transactions in it, and then the collection of transactions was encoded as a standard database on our local machine.
This is definitely a very strange way of retrieving the blockchain for research purposes. Couldn't they have simply issued RPC calls to the regular bitcoind client after downloading the blockchain via the built-in peer-to-peer mechanism?
- carbocation 13y agoYes. In fact, I'm sure (formally, "I suspect") that what you are describing is exactly how blockchain.info gets its data in the first place.
- runeks 13y agoNo. Blockchain.info indexes the block chain by address. bitcoind indexes the block chain by transaction IDs (if you enable it). You'd still have to index the data you get from RPC calls by address if you want to track the movement of certain coins. It's a lot easier to just get it from blockchain.info.