Proposal: DatDot - p2p hosting network built with Substrate
DatDot team is building an autonomous hosting network for p2p data systems. Think of it as a Filecoin but for Hypercore protocol, built with Substrate.
Hypercore protocol (started in 2013) is a distributed ledger technology and set of data syncronization primitives with focus on immutable history.
Due to the P2P nature it's hard to have guarantees about the availability of p2p data systems. If we want to keep data available and up to date we have to keep our computer running or we have to rent a server.
Challenges: What if you want other people to help you host your data?
- what incentives do they have?
- how to find them and trust them?
- how to verify they are ‘seeding’ your data?
Solution: A bridge between Hypercore Protocol and Substrate with a built-in incentive model which manages relationships between:
- hypercore creators/publishers and
- hosters (who keep data available/host the data)
Our team has been active in the blockchain space for many years. Nina Breznik and Alexander Praetorius have been contractors with the Ethereum Foundation for about 3 years and Joshua Mir has been working with Parity tech and is an active member of the web3 ecosystem, especially Kusama network.
Comments (8)
Side note: you should avoid publishing editable hackmd link. Anyone can hijack the content. Questions: Are you asking for 100% of the development cost? Can you provide some justification of why should treasury pay for 100% (or whatever x%) of the cost? If the work benefits everyone in the Polkadot ecosystem, 100% is good, if only a single project that is not Polkadot, that means 0% to me. I will image most of the proposals will fall somewhere in-between. Can you provide more details on how this proposal benefits to the Polkadot Network? Since you are building open source pallet that aims to be used by other projects, it will be helpful if there are some external endorsements as proof of usefulness.
it's honestly quite clear that the storage layer that projects in the broader cryptocurrency ecosystem have settled on is IPFS - but IPFS is not universally suitable for every usecase - hypercore/dat is designed for usecases that are close-to-realtime (ie, streaming) and mutable (append only, though, so still accountable).
Hypercores provide these properties on their own, without the need for DatDot, but similarly to the IPFS ecosystem, you need goodwill seeders or incentivised pinners to host your content to have any degree of persistence.
We want to skip the "pinata" (and other centralized pinning service) era for people to go straight to the "automated and cross-chain" era that the IPFS ecosystem is currently working towards as well, and give dapp developers a more dynamic storage option than IPFS does, with a persistence layer that has been designed for the underlying storage layer.
I can't immediately give you this - perhaps Nina could drop some lines about projects in the Dat Community that are waiting on DatDot or something similar, because there are no projects in the Polkadot ecosystem currently using the hypercore stack to my knowledge.
We had been in early discussions with projects like Joystream, Subsocial, and Kodadot - but no concrete signals of adoption other than an Enum value to select hyper as your storage backend in Subsocial (with no frontend support) - it's a chicken and egg problem though, because projects are looking for the storage persistence/pinning layer (Think filecoin, Crust, Permaweb) and using the availability of those services to decide on their storage layer (IPFS, Arweave).
tl;dr it's quite likely that nobody will use hypercores in the Polkadot ecosystem until DatDot exists, but we argue that hypercores enable unique and useful functionality over the existing set of tools available to the ecosystem, and can therefore expand the set of experiences that can be created within the Polkadot ecosystem.