Starlog: Aiming for the Stars
The full article was originally published by André Fialho on Medium. Read the full article here.
A Metadata Protocol Based on Substrate for the Next Generation of the Web
Ever since we started working with distributed file storage systems (such as IPFS), we faced the challenge that there were no reliable solutions for distributed metadata. However, the potential for this type of information is huge.
First, it provides a wide range of use-cases: powering distributed search engines, enabling decentralized namespaces, providing clear ownership structures, or helping with data curation. Second, it provides a potential solution for GDPR compliance and the issues raised by the EU Copyright Directive. Finally, validated distributed metadata is a key requirement for the development of different smart contracts, such as digital marketplaces. This challenge led us to think about how to implement such a distributed metadata system.
Since we believe the next generation of web needs to be free and easily accessible, we implemented a prototype using, as a first step, the distributed ledger IOTA as a database layer. Transactions on IOTA are feeless and do not require the ownership of anything network specific. Our IOTA based proof of concept dweb.page showed that a fully distributed search was possible. However, IOTA currently doesn’t support the implementation of smart contracts or a governance structure. That was why we started to look for an additional network layer besides IOTA.
This led us to the design of a “metadata network” consisting of a blockchain / smart contract layer called Starlog, an immutable and distributed database layer called Captain’s Log and a distributed storage layer called Starspace. By splitting the technology stack into these different layers, each layer can be adapted according to their key requirements. To be more specific, while the Starlog layer has a high level of distribution and only handles the most important information, the Captain’s Log can take care of the metadata more efficiently and has lower distribution requirements. It’s important to point out, however, that every layer could use different technologies, and that the only key requirement is that these technologies are able to “talk to each” other in the future.