search

filecoin-project / devgrants

👟 Apply for a Filecoin devgrant. Help build the Filecoin ecosystem!
Other
369 stars 307 forks source link

Open Grant Proposal: Part 2: `Improving Socio-technical Resilience in IPFS' #1139

Closed kelsien closed 1 year ago

kelsien commented 1 year ago

Open Grant Proposal: Part 2: `Improving Socio-technical Resilience in IPFS'

Proposal Category: 'research' / devtools-libraries

Proposer: `kelsien'

Technical Sponsor: `Dietrich Ayala (PL)'

Do you agree to open source all work you do on behalf of this RFP and dual-license under MIT and APACHE2 licenses?: 'Yes'

Research Phase 2 Project Description & Research Problem: Data Governance Patterns for Resilient Infrastructure (Data DAOs and More)

This is a proposal for a follow on grant, for a research phase 2 (see below for background and context to the research).

Research Phase 2 builds on the findings from Stage 1 to investigate theoretical and applied models to enhance resilience for IPFS users to forge governance norms and processes that function well (or don’t function, and why).

Resilience, in this context, refers to adaptability and transformability in change or crisis (Tantri & Amir, 2019). In the case of IPFS specifically, this was found to be data integrity in the context of the person using this technology, and in line with their goals.

This scope of work proposes the observation, analysis, and communication of institutional infrastructure for governance models of data content addressed on IPFS. Institutional infrastructures are the political, legal and cultural institutions that form the backdrop for economic activity and governance by enabling or constraining operations that organize and configure societal relations (Hinings, et. al., 2017). For example, the assemblages of people, processes, social norms, and relationships in how data is content addressed, stored, and maintained.

As the lead researcher, I will articulate relevant existing governance models and develop novel approaches where necessary to draw on collective, participatory governance frameworks and organisational design principles to surface ‘best practice’ data governance models. This includes extending my field pioneering research on Data DAOs and data trust models. The data governance pattern library toolkit will be tested in re-interviews with key IPFS implementer stakeholders and shared publicly through IPFS/Filecoin channels, my blog and social channels, and a public facing web page.

Deliverables

Output 1: Use case dissemination to IPFS and Filecoin developers and ecosystem team members, delivered in the form of an online, 2-hour workshop with relevant FF, PL, and implementers from research phase 1 to integrate previous learnings with relevant PL/Filecoin/IPFS team members and test ideas for phase 2 in a workshop format.

Output 2: Resilience enhancement tooling / risk mitigation targeted towards IPFS users, delivered in the form of a ‘Self-Infrastructuring Data Governance Pattern Language Library’ tool that provides a resource for those implementing IPFS on various storage options and how these connect to real-world case studies to inform their infrastructural set-up and improve resilience. The patterns themselves will be presented as an accessible, diagram format that represents people (and IPFS nodes), storage infrastructure, storage providers (where relevant) and linkages (see format inspiration from Community Rule, which is a project of Metagov.org, of which I am a member). The library will be hosted on a public webpage.

Output 3: End user engagement and testing of the tool via, a blog post and social media sharing strategy to communicate and disseminate the toolkit and seek feedback. The target audience for engagement is research participants, IPFS Camp attendees as IPFS implementers, and other relevant stakeholders as identified in conjunction with the PL ecosystem team to test the toolkit and refine the pattern language library. This is incredibly important to grounding the research in real world applications, and connecting with user-communities is a practice I have ample experience in as an ethnographer.

Outcome 4: Better understand how existing governance frameworks (e.g. DAOs, data trusts, and data unions) can be translated and applied to technical requirements and specs, through a long form case study write up.

Value of the Research

This research provides a practical, evidence backed tool to help users of IPFS design, implement, and maintain their storage infrastructure, as well as a detailed use case to continue to develop the research area of data DAOs as a model for decentralised data governance. The goal is to improve resilience for implementers and end users, as data integrity in line with their preferences and local context through non-technical means. The tool also provides a feedback loop to inform product development, UX, communications, and product-user fit, through better understand the contexts and needs of those looking for decentralised content addressing and storage solutions and help implementers to recognise that resilience comes through the rules that they structure for engaging with IPFS and storage infrastructure.

The value of the research is to:

  1. Generate a better understanding for users on the appropriate IPFS and a data storage implementation in their local context, bridging the gap between technical know-how and governance of infratructure for resilience, and how these tools have been effectively employed by others.
  2. Assist IPFS users in getting a better understanding of the uses and limitations of IPFS network in their local context to improve thier resilience, as well as the tools and applications available to assist them in achieving resilience in their use of IPFS and inform the language used to communicate IPFS.
  3. Produce artefacts (i.e design tool + written use cases)to inform technical implementations (e.g. tooling) and standards development surrounding decentralised storage infrastructure (for example, at W3C and IETF).
  4. Provide a grounded evidence base to inform decision-making by application developers, service providers, and governance process designers building on top of or alongside IPFS protocol, both within PL/IPFS/Filecoin and outside of these organisations in relation to decentralised data governance.

Research Process to Address the Problem

The research methodology is qualitative mixed methods. It will include ethnography and design anthropology. Ethnography is the study and communication of people. Ethnographers search for patterns in the lived human experiences by carefully observing and participating in the lives of those participating in the study (Angrosino, 2008). The data collection methods used are ethnographic interviews and digital ethnography (observation of online channels). The advantage of ethnography is that it can provide rich, cultural insights that are otherwise invisible to offer a feedback loop on the use of digital infrastructure (Star, 1999). Design anthropology is the mobilisation of theory, methodology, and practice to create interventions. It has been used in the development of emerging technologies to create better human outcomes (Pink, et. al., 2020). Qualitative methods are an important compliment to quantitative protocol design and development, to capture and understand design and user data that is not readily available or visible. The University ethics approval for this research project is under the PhD project “Resilience in Decentralised Technologies” (see Appendix A). I have experience with decentralised governance research and design alongside engineers and data scientists at BlockScience.

The guiding research question that will guide this scope of work is: What storage options and data governance patterns are available for people implementing IPFS in their local context?

The research sub-questions are:

So far in this study, I have defined the concept of resilience in the context of sociotechnical systems, by employing the concept of stakeholder, context, threat, and vulnerability. The aim of phase 2 of the research is to address vulnerabilities in institutional infrastructures that make or break the resilience of use of IPFS network through the data governance pattern library.

Assumptions and limitations: a. Desktop research methods will be adopted, with ethnographic interviews where access is appropriate and possible, assuming access to mutually agreed user communities will be possible. b. Conducting research interviews with stakeholders identified will not cause risk to either the researcher or research participants. If this is deemed a risk under research ethics, the researcher reserves the right to not pursue an interview. c. Relevant stakeholders will be available in a timely manner for research interviews. d. Research will be conducted remotely, via online ethnographic interviews and digital ethnography. e. The designated Protocol Labs contacts will be available in a timely manner for research interviews, feedback, and project logistics. f. Project timelines are estimates only.

Development Roadmap

  1. Initial workshop
  2. Design tool V1
  3. Communication of tool, user engagement and request for feedback, analysis and reporting.
  4. Data DAO case study write up.

Research Deliverables:

Milestone based.

a. 1 x 2 hour workshop with relevant PL team members to workshop initial findings/requirements, share design tool ideas, and gather feedback (e.g. ecosystem team, ResnetLabs, governance team, PL research fellows, PL implementers). b. 1 x Design tool prototype. Engage with implementers for feedback. Match design archetypes to case study examples. c. 1 x Blog to communicate tool, feedback engagement, and live website. d. 1 x research paper length case study on actioning the concept of data DAOs

Proposed timelines:

February, 2022. Workshop Deliverable a.

March – May, 2023. Literature review and theory development. Design prototype. Develop and implement communication strategy. Deliverable b.

June – August, 2023. Engagement with IPFS implementers, collate feedback, iterate tool.

Deliverable c. September - November, 2023.

Review and debrief. Identify further areas for research and development, case studies, etc. Engage in any other case study write ups or open research and conference initiatives to communicate results as appropriate (e.g. IPFS Camp).

A 12-month project timeline for the deliverables outlines is anticipated, with considerations for end-of-year breaks in accessing research participants and pending the assumptions and limitations listed below. The timeline and scope of the project may be reviewed upon agreement of PL and the Researcher to include other activities, such as involvement in internal stakeholders, standards body working groups, or other communication, presentation, and research dissemination efforts.

Total Budget Requested

1 x PhD Fellowship equivalent of $60,000 in USD or FIL. Maintenance and Upgrade Plans

This is Part 2 of the research plan, as intended upon commencement of Part 1. My intent is that quantitative research collaboration opportunities arise out of this project (especially network resilience mapping), as well as further engagement with other qualitative researchers on IPFS and Filecoin (especially governance research - for example, that initiatives like IPFS “DINPS” workshop could have a social science track, as well as technical research tracks).

Team

Team Members:

Kelsie Nabben Github: kelsien LinkedIn Profile: https://www.linkedin.com/in/kelsien/ Website: https://rmit.academia.edu/KelsieNabben/Papers Blog: https://kelsienabben.substack.com/

Relevant Experience

Researcher Profile: Kelsie Nabben, Researcher/Project Manager, RMIT University Blockchain Innovation Hub / Digital Ethnography Research Centre / Centre of Excellence for Automated Decision-Making & Society. Kelsie’s research focuses on resilience in socio-technical systems. As an ethnographic researcher, she is interested in understanding the contexts, uses, and social outcomes of decentralised technologies and the governance of these systems to ensure resilience. She is also actively engaged with industry engineering firms and decentralised governance research groups with BlockScience, Metagov, and DAO Research Collective. The findings of the related Part 1: ‘Socio-technical resilience in IPFS’ 2021 research grant resulted in a Protocol Labs – IEEE peer-reviewed paper which was presented at DINPS ’22 workshop and published in IEEE.

Team code repositories

See above links.

Additional Information: Background & Context (Research Phase 1)

The first phase of the research grant consisted of qualitative research on the topic of ‘resilience’ in various contexts of threat in real-world usage of IPFS and Filecoin network. This led to the analysis and write up of four use cases (on NFTs, Matters News, Omnilingo, and Starling Lab) that were publicly communicated via blog post, peer reviewed paper, and in-person workshops and talks at DINPS, Funding the Commons, and IPFS Camp.

Each deliverable of the funding proposal was met on time, with the outcomes of the research exceeding expectations, in terms of:

• General interest in the research across IPFS/ PL / Filecoin in better understanding people’s end user experience and generation of a channel in Filecoin Slack. • Documenting some crucial IPFS use cases for the first time. These can be used as case studies to inform technology standards development with industry bodies. • Engaging with existing users that have received PL grants to better understand their experience of the technology (user feedback). • The ecosystem responded by expanding initiatives like DINPs research conference to include these research findings and engage with other researchers. • Two talks and one technical track workshop (in Yiannis’ session) at FIL Lisbon/IPFS Camp.

The findings of the initial phase of research pertain to how people’s coordination around technical infrastructure impacts resilience (aka. the importance of ‘institutional infrastructure’). This includes:

  1. IPFS solves a genuine user need for content addressing that is not based on HTTPs across a number of use cases where threats to data persistence can be ‘active’ (such as nation-state level censorship) or ‘passive’ (such as data loss),
  2. A unique moment of infrastructural vulnerability occurs in real-world use cases when content addressing is connected to storage. In practice, people implement storage in a number of bespoke ways, with only some using Filecoin (and noting the infrastructure and tooling is still under development).
  3. Further research and development of theoretical and applied models on the way that people coordinate around a shared infrastructure (‘institutional infrastructure’, including tools, systems, processes, and governance), as essential components for resilience of IPFS and Filecoin in practice.
ErinOCon commented 1 year ago

Hi @kelsien, thank you for your proposal! This grant has been approved. Can you confirm the preferred email for discussing next steps?

kelsien commented 1 year ago

Amazing! DMing you

ErinOCon commented 1 year ago

Hi @kelsien, thank you for providing your email address. Looking forward to hearing from you!

kelsien commented 1 year ago

Research update: March, 2023

Progress Update: The project is running according to timelines. Deliverables have been slightly re-ordered in relation to the project plan as the pattern library (or 'tool') has been developed ahead of the workshop to best communicate research findings and receive input and dissemination.

The decentralised data governance pattern library is available here. I have developed it on 'portrait', a beta tool for building websites on IPFS.

The next step of the project is to list stakeholders and set a date suitable with them for the workshop to share the key research findings with developers and other FF and PL stakeholders, gather further feedback from interviewees and IPFS users, and further inform the research dissemination plan.

I have also engaged an experienced decentralised technology researcher to support on development of some of the written case studies and management of new contributions to the pattern library once it is shared out.

New directions arising from the research are:

Open questions arising from the research project so far are:

I am meeting with the project point person from within PL next week to report progress and plan the workshop.

kelsien commented 1 year ago

Research update May 2023:

kelsien commented 1 year ago

Research update August 2023:

  1. Pattern library complete and published https://github.com/kelsien/datagovernancepatterns and website https://beta.portrait.gg/0x1018c3A71997D2171E915ea89f704d9a192823EF

  2. invite only workshop scheduled for August 16 and well subscribed with industry practitioners (IPFS implementers), researchers, and internet infrastructure experts. Agenda designed in conjunction with project PoC at PL.

  3. Milestone D is underway, and will include a general introduction to the concept of data DAOs, an in depth case studies, and a technical pre-PoC on how data DAOs apply in the context of Filecoin. ETA is according to pre-established project timelines.

kelsien commented 1 year ago
  1. Workshop conducted and participants invited to Filecoin slack to continue the conversation.

Blog (deliverable c) has been published. See: https://kelsienabben.substack.com/p/announcing-the-decentralised-data /

https://kelsienabben.medium.com/announcing-the-decentralised-data-governance-pattern-library-exploring-social-arrangements-cd68deeea22b /

https://kelsiemvn.mirror.xyz/IM2LZ1X2sl6pjyCW2cYT2_MgPOopASnQLCNJ_1Lb3UE /

https://twitter.com/kelsiemvn/status/1691965515928109258

kelsien commented 8 months ago

Research update December, 2023:

Deliverable d (1 x research paper length case study on actioning the concept of data DAOs) is in the final stages of production (ETA end of December).

This has been expanded to 1x full academic paper (awaiting final feedback from research participants to ensure it complies with research ethics) as sell as one long-form blog post for implementers on the concept (with a more pragmatic, technical focus - that has been co-authored with engineers at BlockScience).

Side note: won a 'Vice Chancellor HDR Research Impact' Award, including for this collaboration, and passed my PhD!

kelsien commented 4 months ago

Project update, April 2024:

Current Milestone = Outcome 4: Better understand how existing governance frameworks (e.g. DAOs, data trusts, and data unions) can be translated and applied to technical requirements and specs, through a long form case study write up.

Progress Report:

Rather than one case study, I have decided to split these deliverables into two for greater impact: (1) A technical paper on 'Preliminary steps in designing a data intermediary', and, (2) An academic paper, including case study, of 'An ethnography of a data intermediary'.

I have engaged a third-party organisation to collaborate with technical co-authors and compensate people's time in data collection for the case study. Although constructive, significant delays in final edits and approvals from all parties involved.

Both outputs have been in full draft since December. My hope is that output 1 will undergo final edit and be released by May 2024 at the latest, and output 2 will be completed and submitted to a publication venue, and released in pre-print format for dissemination and engagement.

kelsien commented 2 months ago

Progress Update: July 2024

TLDR: More technical paper is out, satisfying the requirements of Outcome 4 (Better understand how existing governance frameworks (e.g. DAOs, data trusts, and data unions) can be translated and applied to technical requirements and specs, through a long form case study write up) and Milestone D ( 1 x research paper length case study on actioning the concept of data DAOs).

See: _Sisson, David; Nabben, Kelsie; Ben-Meir, Ilan; and Zargham, Michael, Data Mesh Architecture: Interoperability, Co-Operation, and Co-Regulation (June 29, 2024). SSRN. https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4880709_

Context:

This Outcome has evolved since the initial proposal. I brought a technical partner on board for domain expertise, and to bring their experience within the Filecoin ecosystem to bear.

  1. Initially, this outcome began with general research stemming from the data governance pattern library on data DAOs and data trust examples, implementation practices, and considerations (SEE BELOW).
  2. I then looked for data DAOs on Filecoin, of which there are very few functional ones (e.g. FVM Data DAO Example: GitHub - rk-rishikesh/DataDAO. A notable exception of TRANSFER Data Trust ((which is a trust, not a DAO - https://github.com/TRANSFERArchive/DATA.TRUST), and has just been announced. I conducted a research call with Kelani in 2023, to whom I am grateful for her time.
  3. I then scoped work with an expert from BlockScience on designing a PoC spec for a data DAO on Filecoin. We determined that the requirements for a data DAO PoC spec on Filecoin were too broad and unclear.
  4. What followed was a research paper, oriented towards a technical audience, that explores the more actionable concept of a data mesh, as a precursor to how this might apply in the Filecoin ecosytem (see Sisson, David; Nabben, Kelsie; Ben-Meir, Ilan; and Zargham, Michael, Data Mesh Architecture: Interoperability, Co-Operation, and Co-Regulation (June 29, 2024). SSRN. https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4880709)
  5. In terms of a more general case study, a significant amount of work went into interviews related to the Superset DAO and data trust (https://www.supersetdao.com/). This was written up but not finalised, in favour of collating already published work on Superset (SEE BELOW), and slated forthcoming papers from trustees involved in the ecosystem.
  6. I am writing a forthcoming general presentation on the role of data intermediaries and accountability in data ecosystems, based on a presentation given on a panel at the Platform Economies Research Network in April 2023, that has been invited for a special collection (noting, I am beholden to academic editorial timelines on this but will share when complete).
kelsien commented 2 months ago

RESEARCH SUMMARY : What are best practices for implementing data DAOs?

Because data DAOs are a new organizational form, there is no playbook for building effective, equitable and sustainable data DAOs (yet). Some best practices might include: • A DAO constitution or charter to clearly define the mission of the data DAO and its initial principles, policies and procedures regarding the data that it manages, as well as explicitly defining the scope of tokenholder governance • Establishing working groups or subDAOs dedicated to data stewardship to maintain data quality and interoperability, and governance stewardship to ensure effective tokenholder governance • Clear and detailed technical documentation to ensure that important knowledge of the data architecture and technical infrastructure is shared beyond the founding members of the DAO, removing barriers to participation by developers and facilitating informed decision-making around technical issues in tokenholder governance • The creation of legal entities for the DAO and potential subDAOs ensure any legal liability does not fall on DAO members as individuals. • A rigorous and cautious approach to tokenomic design that seeks to keep volatility to a minimum and preempts any perverse incentives that would be detrimental to the sustainability of the data DAO • Exploring more streamlined approaches to governance that reduce the time required to reach consensus and action a decision such as reducing the number of voting stages or establishing a delegation system that uses constrained delegation; preempting risks of governance manipulation or attacks in the design of the voting system • Explore both financial and non-financial incentives to bootstrap the data DAO’s network, potentially prioritizing data supply over demand in initial stages of the DAOs roadmap

Examples of data DAOs using decentralized data storage and governance technologies Lagrange DAO Lagrange DAO is a data DAO that offers data sharing and analytics infrastructure (“Lagrange Spaces) for research collaboration and a decentralized computing power renting network that allows community members to rent out their computing power to researchers and users. The DAO uses a token-based incentive system that rewards contributions to the DAO. These tokens can be used to pay for data services, rent computing power or participate in DAO governance. https://docs.lagrangedao.org/lagrange-dao/

Further resource on data DAOs Kangaroo, a DataDAO starter kit: https://github.com/kangaroo-data-dao/ An FVM compatible base data DAO contract for implementing a data DAO on Filecoin: https://github.com/rk-rishikesh/DataDAO From D > interesting hackathon project - Kangaroo Data DAO starter kit • https://devfolio.co/projects/fevm-data-dao-starter-kit-kangaroo-2f3d • https://ethglobal.com/showcase/kangaroo-fevm-data-dao-starter-kit-67ic7 • GH: https://github.com/kangaroo-data-dao/kangaroo-data-dao • live: https://kangaroo-ten.vercel.app/

We Need More Control Over Our Own User Data, an essay by Superset governance board member Amber Case https://medium.com/block-science/we-need-more-control-over-our-own-user-data-43f267a817f5

Decentralised Autonomous Organisations (DAOs) as Data Trusts: A General-purpose Data Governance Framework for Decentralised Data Ownership, Storage, and Utilisation, a working paper by Kelsie Nabben https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4009205

What is a Data Trust? By the Open Data Institute https://theodi.org/article/what-is-a-data-trust/

DeFi & DeSci Data DAO examples: https://nevermined.io/projects/ https://nevermined.io/project/defi-data/ https://nevermined.io/project/hairdao/ https://stateborn.org/about

kelsien commented 2 months ago

RESEARCH SUMMARY: Resources on Superset DAO and Case Study

Case, A. (2023). We Need More Control Over Our Own User Data. Superset. April 4. Available online: https://medium.com/supersetdao/we-need-more-control-over-our-own-user-data-43f267a817f5. Accessed 4 December, 2023.

Alston, Eric., Ben Bartlett, Amber Case, Andrew Peek, Michael Zargham, and O’Neil Risk Consulting and Algorithmic Auditing, Inc (2023). “DECLARATION OF TRUST ESTABLISHING THE SUPERSET TRUST”. 2 May, 2023. pp. 1-34. Available online: https://uploads-ssl.webflow.com/639b7ea7130f2821fe42d8fc/6408bd215b28260b8c50795f_Superset_Trust_Declaration.pdf. Accessed November 1, 2023.

Superset. (2023). “How Superset Helps Members.” Gitbook. Last updated 04/6/23. Retrieved November 10, 2023, from https://deep-work.gitbook.io/superset/welcome-to-superset/about-superset/how-superset-helps-members.

ORCAA. (n.d). ORCAA. Available online: https://orcaarisk.com/. Accessed October 31, 2023.

Ruhaak, Anouk. (2020). “Data trusts in Germany and under the GDPR”. Algorithm Watch. Available online: https://automatingsociety.algorithmwatch.org/. Accessed 1 November, 2023.

Sylvie Delacroix, Neil D Lawrence, Bottom-up data Trusts: disturbing the ‘one size fits all’ approach to data governance, International Data Privacy Law, Volume 9, Issue 4, November 2019, Pages 236–252, https://doi.org/10.1093/idpl/ipz014.

Alston, E., Killian, S., David, G. Killswitch. https://summerofprotocols.com/research/killswitch-protocols

kelsien commented 2 months ago

Next Steps:

  1. Implement final comments on paper from author team
  2. Kelsie to post blog and share on socials + amplify via networks (including Filecoin)
  3. Kelsie to notify grants team (no outstanding items)
  4. Kelsie to invoice (final invoice)

Timelines: ASAP, pending co-author responses. I would hope by July 30 at the latest.

kelsien commented 1 month ago

31 July update: paper is live at: Sisson, David and Nabben, Kelsie and Ben-Meir, Ilan and Zargham, Michael, Data Mesh Architecture: Interoperability, Co-Operation, and Co-Regulation (July 24, 2024). Available at SSRN: https://ssrn.com/abstract=4880709 or http://dx.doi.org/10.2139/ssrn.4880709

kelsien commented 3 weeks ago

11 August update:

Blog post to dot point summary and promote paper is live (https://kelsienabben.substack.com/p/data-intermediary-networks + https://medium.com/block-science/data-intermediary-networks-3d81209e8be6).

Promoted on social media (https://twitter.com/kelsiemvn/status/1820966374585593872 + https://twitter.com/block_science/status/1821021382677082428 + https://twitter.com/BlockchainGov/status/1821464540950811108).

Awaiting support from Filecoin mar/comms to socialise the core contribution of data intermediary networks > an actionable architecture to hone the concept of 'data DAOs' in the IPFS + Filecoin ecosystem.

Scope for next steps includes running general workshops on this concept and working with builder teams to develop and implement a PoC