Published on: 2013-02-11T19:39:03+00:00
The email thread discusses the use of blockchain for establishing identities and as a "root CA" to sign arbitrary certificates. The author suggests that signing the root in the DNS tree requires more effort than building a CA off the bitcoin blockchain. To associate identifiers for payment addresses, the writer suggests putting those in DNSSEC signed records in the DNS. For routing around x.509 CAs, participation in the DANE working group in the IETF is recommended. Timo Hanke argues that DNS has lower security demands compared to payment IDs for Bitcoin. He believes that DNS data can be in a chain with a lower hash rate than Bitcoin's hash rate while payment IDs need to be in a chain with the same hash rate. However, the person he is communicating with may not be well-informed about what Namecoin does - it is a multiple namespace key-value store that can have exactly the same Proof of Work hash power as a non-parasitic system. Namecoin chose a model that allowed it to survive even if Bitcoin failed, but this is not necessary for others to do. The writer strongly recommends externalizing the costs of services without taking resources from Bitcoin currency users without their consent.In a discussion on the Bitcoin development mailing list, Timo Hanke proposed to establish an identity for a payment protocol using a new namespace rather than tying names to data like Namecoin does. Hanke suggests that DNS has a lower demand than payment ids have, and therefore, DNS data can be in a chain with a hashrate lower than Bitcoin's hashrate. Payment ids meant for Bitcoin, on the other hand, must be in a chain with equal hashrate. Hanke also discusses merging the new chain with Bitcoin's blockchain. However, he questions whether this would still be considered an alt-chain or a more efficient version of what he proposed.Peter Todd, another developer on the mailing list, suggested that creating a new system adds to the confusion for users who cannot memorize unique numbers and prefer non-unique names. Todd recommends working on Namecoin to make it more usable while adding some PKI to it using the same domain names. However, Hanke argues that Namecoin is unsuitable for his purpose and proposes using a payment base address, which he sees as a more universal identifier than a domain name. Additionally, Hanke explains the pay-to-contract principle, which means authenticated communication is not necessary when paying, so funds are never lost even if ordered from a wrong website.The developers also discuss where to store alt-chain block header hashes. Hanke initially thought of putting them in the UTXO to enable a future dedicated "hardware wallet" to validate against it. However, Todd points out that storing all the data in the UTXO would mean users need to have the entire blockchain available, which is not desirable. Ultimately, they agree that storing the alt-chain block header hashes outside the UTXO would be better.In this email conversation, Timo Hanke justifies why he does not use Namecoin or any other alt-chain to establish an identity for a payment protocol. He explains that his incoming payments will land on addresses that are derived from this identity, making it as important (security-wise) as anything else involved in the bitcoin protocol. Therefore, he would not want to have payment-ids rely on anything less than bitcoin's own blockchain. However, the person replying to him argues that creating another system is just adding to the confusion of users, who can be tricked into thinking some name snatched up by scammers in some little-used PKI system is who they are supposed to be communicating with. The person suggests that Timo should work on Namecoin and make it more usable, then add some PKI to it using the same domain names so that when they see a PKI certificate for "foo", they know it must be the same "foo" website they just visited and the same "foo at foo" they just emailed. They also suggest that Timo could register a domain name in an alt-chain, then "mine" the block by creating a suitable transaction, make sure it's the biggest fee, wait a few confirmations, and they're good to go with the same level of security as Bitcoin proper. Multiple parties can also collaborate to create the transaction, each providing and signing for an input for their portion of the total fee. Moreover, they propose a "two-step sacrifice" protocol to solve the problem of miners getting to keep the fee, making it difficult to get a block mined, or to be sure a block won't be undone later. Finally, they discuss the cost of storing the UTXO set in memory, saying that it would be better to store it in the spent outputs.
Updated on: 2023-08-01T04:25:00.885958+00:00