Most of the ideas I tumbled across are based on recipient's address filtering.
This concepts seems to me nothing but absurd, since whatever filter is implemented, there will always be a way to fool such filtering process, and store data.
Here, I propose a method that:
- Eliminates arbitrary data-storage in the blockchain
- Eliminates lost money due to mistyped addresses (if ever anyone types an address)
The solution is as simple as this, both payer and receiver must sign the transfer. This could've brought up the double-spending issue again,
but here is how to solve it.
Bob wants to pay Alice
Bob sends transfer to Alice's address
Such transfer is confirmed in the blockchain
After such confirmation, Alice must sign such transfer within the next N block confirmations (e.g by re-transfering the money to another (her's) address, or by creating a new type of packet "transfer confirm")
if after N blocks no-one confirmed the receipt of the money, bob is allowed to spend it again. Such transaction can then be removed.
How does this attenuate the "dark" data storage?
The person that wants to store data, sending to a fake address (which contains the given data) needs to have the private key of that address - > which is pretty complicated.
Nevertheless, it is still possible (and will always be) to store data using a vanity address...
How does this solve "mistyped" addresses?
The receiver cannot sign the receipt, and as such the money goes back to the sender
What do you think about this proposal?
Edit: made some makeup on text
Even banks are more practical than that