Orcfax: trustworthy Cardano oracles

More background on the problem:
Blockchains are said to be "trustless" thanks to cryptographic proofs. However, smart contract developers need access to *trustworthy* information external to the blockchain to serve as inputs to their application logic. For example, a change in the BTC/ADA price may trigger a limit order on a DeFi platform, the final score in a soccer match may trigger a payout in a betting Dapp, or an extreme weather event may trigger a crop insurance claim.

Oracles are the blockchain primitives that provide this real-world information to on-chain smart contracts. The data they provide must be authentic as it will have significant financial consequences. However, there is a risk that the critical role of oracles in the Cardano smart contract ecosystem will be overlooked or downplayed in face of the (justified) Goguen era hype and the flurry of pre-launch project announcements.

The small crop of existing oracle solutions, such as the Ethereum big fish Chainlink, do a good job articulating the "oracle problem" but when you dig deeper you see that they often conflate information security with information provenance. If you look past the slick ICO marketing and whitepapers you are really being asked to trust their technical decisions and collaborators rather than an open specification and easy-to-audit architectures. Even with sound security protocols, the opportunities for manipulating oracle information somewhere in their provenance pipelines remain ripe without the solution being proposed by the Orcfax project.

Our solution:

A standards-based specification and open source software that *explicitly* expose and archive the flow of information from real-world data sources (e.g. APIs, open datasets, reports), through oracle architecture stacks, and onto the Cardano blockchain.

The project is called "Orcfax" as a play on words combining "oracle" and "facts." A Cardano-branded orca is our logo.

We will publish the Orcfax Open Oracle Specification and develop an open-source, Cardano-native reference implementation. This software will act as a pragmatic proof-of-concept for the specification as well as provide Cardano developers and oracle providers with a free and open source tool to deploy their own Orcfax-based solution.

To draft the specification we will assess and apply proven industry standards that were developed over many years of open collaboration by communities of domain experts. These include:

• W3C PROV-O: the definitive specification for expressing information provenance. See https://www.w3.org/TR/2013/REC-prov-dm-20130430/
• W3C Verifiable Claims: the definitive specification for third-party verification of online information. https://www.w3.org/TR/vc-data-model/
• ISO 17068: the definitive specification for maintaining trustworthy repositories of authentic information. See https://tinyurl.com/39acbewk
• ISO 27001: the definitive specification for maintaining an information security management system. See https://www.iso.org/standard/54534.html
• NIST Zero Trust Architecture: the definitive report on best-practice security architecture that focuses on constant authentication and authorization instead of network-based privileges. https://nvlpubs.nist.gov/nistpubs/SpecialPublications/NIST.SP.800-207.pdf

We will also reference the nascent body of academic research on blockchain recordkeeping, particularly in the field of archival science and research data management. See for example:

• Building Decentralized Trust (2021) https://www.springer.com/gp/book/9783030544133
• A Blockchain-Based Notarization Service for Biomedical Knowledge Retrieval (2018) https://www.sciencedirect.com/science/article/pii/S2001037018300400
• Trusting Records in the Cloud (2019) https://www.facetpublishing.co.uk/page/detail/?k=9781783304028

Additionally, we will analyze ubiquitous tools for managing information integrity and re-usability that offer an opportunity to bridge trustworthy Cardano data with the wider web domain where most of the online world will continue to conduct its day-to-day affairs. These include Git, IPFS, JSON-LD, Schema.org, and Wikidata.

The reference implementation software will be a working proof of the Orcfax Open Oracle Specification, both of which will be published to our Github repository under open source licenses.

The reference implementation's technical architecture will leverage Cardano's unique on-chain metadata blocks and copy best-of-breed oracle design patterns such as those demonstrated by Chainlink, Web3 API, and Charli3. We are most excited about applying the Oracle Pools design, first publicized by Emurgo, which normalizes multiple data sources and leverages Cardano's unique Extended UTXO design.

Defining success:

After one month:

• The first draft of the Orcfax Open Oracle Specification has been circulated for community review and feedback.

After three months:

• Version 1.0 of the Orcfax Open Oracle Specification has been published.
• The Minimal Viable Product (MVP) of the Orcfax reference implementation has been deployed on the Cardano blockchain.
• Community consultation with Cardano smart contract developers has begun to define the most in-demand oracle endpoints for the first wave of Orcfax oracle deployments: e.g. cryptocurrency price data, sporting event data, weather data, real estate data, public health data, etc.

After six months:

• The Orcfax Open Oracle Specification has been updated based on the practical knowledge gained from the MVP implementation.
• The beta version of the Orcfax reference implementation has been published and has been deployed by one oracle operator external to the Orcfax team.
• The Orcfax project has deployed its own oracle endpoint(s) which has been tested by at least three smart contract developers external to the Orcfax team.

After nine months:

• An Orcfax sustainability plan will be circulated for community feedback. This may include oracle transaction fees, more Project Catalyst funding requests, an Initial Stakepool Offering (ISO) offering, a Decentralized Autonomous Organization (DAO) or some combination of these. We are not putting the cart before the horse. Orcfax first has to demonstrate that it is an intellectually and technically sound solution and that it has garnered community interest. The primary objective of the sustainability plan is to propose the most cost-effective option that will ensure the long-term availability of affordable, trustworthy, real-world information on the Cardano blockchain.

After twelve months:

• The Orcfax sustainability plan has been implemented.
• Orcfax oracles are providing actionable, real-world information to ten+ in-production Cardano smart contracts that hold and distribute ADA.
• The Orcfax Open Oracle Specification is pursuing a formal standardization process (e.g. CIP, IEEE, W3C, ISO).

Budget:
The requested Fund 6 budget is for wages and server expenses for the first three months of the Orcfax project. This will deliver the first version of the Orcfax Open Oracle Specification and the reference implementation MVP:

• 1 x Senior Architect (Peter Van Garderen) x 20 hours per week x $75 USD/hr x 12 weeks = $18,000
• 1 x Systems engineer (TBD) x 20 hours per week x $75 USD/hr x 12 weeks = $18,000
• 1 x Software developer (TBD) x 20 hours per week x $75 USD/hr x 8 weeks = $12,000
• Server and transaction fees: $500 x 3 months = $1,500

The $75/hr rate is on the mid-salary scale for senior professionals in the digital archiving and software engineering field (in the Western hemisphere). All three project roles include time allocated for project management and community engagement tasks.

The team members will provide their own computing equipment. Hosting for the project website and code repositories are provided free of charge via Github. Community outreach will be done via (free) Twitter and YouTube accounts as well as the various Project Catalyst communication channels.

Division of responsibility:
The senior architect and systems engineer both have software engineering skills. They will collaborate in the first month using Agile software development methodology to define and validate the technical requirements for the MVP. While the Open Oracle Specification will be based on the industry standards listed above, it will also be informed by the pragmatic technical requirements identified during this initial analysis stage.

Once the technical stack has been selected in the first month, a software engineer with the relevant language skills will be added to the team to help complete the MVP.

The systems engineer will take responsibility for maintaining and securing the off-chain oracle server infrastructure and Cardano node communications.

The senior architect will be the primary author of the specification as well as the software development supervisor.

Licensing:
The source code for the Orcfax reference implementation will be published under a free and open-source MIT license.

The Orcfax Open Oracle Specification will be published under a Creative Commons Attribution-NoDerivatives 4.0 International (CC BY-ND 4.0) license.

This will make the specification free to use and re-distribute in any format. However, the "No Derivatives" clause means that altered versions cannot be published. This is to avoid the cherry-picking of specific clauses without respecting the full intent of the specification. That would dilute its reputation and validity. Once it's published, a formal revision process for the Open Oracle Specification will be announced.

Who we are:
The Orcfax project is led by Peter Van Garderen, a professional archivist and world renowned expert in the field of digital archiving. See https://vangarderen.net

Dating back to medieval times, the archival profession has maintained standards and practices for protecting and demonstrating the authenticity of documents and information objects. Peter is a pioneer in developing free and open-source software systems that implement these standards. He is the founder of Artefactual Systems, a bootstrapped FOSS service company that provides development and support services for the hundreds of institutions worldwide that have deployed the archiving solutions he developed. See http://archivematica.org and https://accesstomemory.org

Peter is also a long-time decentralization advocate and researcher. He spoke on applying archival principles to blockchain technology at the first Decentralized Web Summit hosted at the Internet Archive in 2016. In the past twelve months he has led development teams that won several Filecoin hackfests with projects that provided a Python interface for Filecoin, archived Open Image datasets onto the Filecoin network, and a tool for migrating Facebook content onto Filecoin.

Peter will draw on his wide network of colleagues, collaborators, and newly found Catalyst contacts to hire an experienced systems engineer and software developer who will help him to refine the Orcfax specification and deliver the Orcfax reference implementation software. He is currently in discussion with several qualified systems engineers for that key role.