lightning-dev

Combined summary - New idea on decentralized identity and truth (Re: Numerifides)

Combined summary - New idea on decentralized identity and truth (Re: Numerifides)

Tyler H has proposed a feature to add Atlas as a punishment mechanism for eltoo channels in Lightning Network.

The feature is being discussed and hashed out on GitHub. Tyler H's idea is that any node can be queried for a mapping, which involves fulfilling a payment request made by the client. This would allow for decentralized, trusted hosting of data mappings, with the cost of longer paths or more queries being higher, making it fair. The database could also be shared among nodes for a price and redundancy. Additionally, queries for non-mapping information like restaurant recommendations could have potential use cases in the future. Feedback is appreciated.Tyler has opened a GitHub pull request to discuss the features and behavior of Atlas. The goal of Numerifides was to create trusted name mappings that are human meaningful. Tyler suggests that every node could be a watchtower, by announcing to the network at large when a channel is breached, thus nodes will likely want to close their channels with the less trustworthy nodes. Tyler proposes adding a feature to Lightning whereby any given node can be queried for a mapping, such as "Give me the IP address for Google.com" and the node can provide any answer one chooses along with fulfilling a Lightning payment request the client provides. Tyler suggests that paying 1 satoshi per query for decentralized, trusted hosting of your data mappings seems fair. Lastly, Tyler suggests that the database could be shared among nodes for a price, where a Lightning node can offer to store data per hour and the person who wishes for redundancy can pay a Lightning invoice and provide the data.The proposal is to add a feature to the Lightning Network where any node can be queried for a mapping along with fulfilling a Lightning payment request. The idea is that important mappings will be paid for, while unimportant ones won't. Longer paths or more queries will cost more, and each query of the mapping will cost 1 satoshi per query for decentralized, trusted hosting of data mappings. The database could be shared among nodes for a price, where a Lightning node can offer to store data per hour and the person who wishes for redundancy can pay a Lightning invoice and provide the data. This data wouldn't have to be encrypted or private since it can be publicly queried. Feedback is appreciated.The idea proposed by Tyler H is to add a feature to Lightning whereby any given node can be queried for a mapping along with fulfilling a Lightning payment request the client provides. The proposal includes adding an RFC and a feature bit to Lightning. This would allow payments for mappings that are considered important, as well as make decentralized, trusted hosting of data mappings possible. Longer paths or more queries will cost more, so paying 1 satoshi per query seems reasonable. The database could be shared among nodes for a price, where a Lightning node can offer to store data per hour and the person who wishes for redundancy can pay a Lightning invoice and provide the data. However, ZmnSCPxj raised concerns about this proposal. The attacker could invent some target audience that it pretends exists, but might not actually exist. Also, non-distributed server-client protocols for storing database information pay in reverse; the querier requests some query, the responder sends the encrypted data and the key to the encrypted data, and the querier pays the invoice and receives the preimage which is the encryption key to the encrypted data. ZmnSCPxj suggested building off the "server-client database" idea to create a special node type called an "advertiser node". Advertiser nodes have a "write" interface that lets advertisers pay to set a mapping, a "read" interface that lets audiences pay to get a mapping, and a "proof" interface that lets anyone query how much the advertiser node owns in its channels. If an advertiser node owns much funds, it probably earned that from many advertisers paying to set mappings, and audiences paying to get mappings. So if the advertising node is inventing its audience, then it will have to lock some of its own funds and not spend it, sacrificing opportunity to invest it elsewhere. However, this is strongly centralizing and thus not recommended.The email thread discusses a proposed use case for Lightning, which involves adding a feature to the network that allows nodes to be queried for a mapping along with fulfilling a Lightning payment request. The proposal suggests that this would allow for decentralized, trusted hosting of data mappings and would only be paid for by those who care about the mappings. Nodes must have inbound Lightning capacity to receive payments, and attackers cannot drain more than they have spent on their channels. To ensure that an advertiser is honest, a proof interface can be used to query how much the advertiser node owns in its channels. The proposal is compared to an alternate idea involving advertiser nodes, but it is deemed too centralizing. The database could also be shared among nodes for a price, where a Lightning node can offer to store data per hour and the person who wishes for redundancy can pay a Lightning invoice and provide the data.Zmn

Discussion History

0
Tyler HOriginal Post
June 7, 2018 02:27 UTC
1
June 7, 2018 03:20 UTC
2
June 7, 2018 03:36 UTC
3
June 7, 2018 04:01 UTC
4
June 7, 2018 08:07 UTC
5
June 7, 2018 08:45 UTC
6
June 10, 2018 22:29 UTC
7
June 10, 2018 22:32 UTC