Here are some open questions we are currently thinking about after the convening.

Please share ideas, experiences, and questions of your own:

Federating ownership and sustainability

• What are effective ways to distribute funding, maintenance, and long-term support for infrastructure to advance publishing and sense-making?

• How can we support multi-institution ownership and sustainability for broader initiatives?

• What networks and consortia are you a part of? Which work well + in what ways? What collaborations have felt particularly impactful, well-timed?
  

• How can we encourage risk-taking and support failure? Bottom-up.
  

• How can we support strategies broader/more roust than immediate siloed tools and problems? Be strategically competitive

Fitting into current infrastructure and institutions

• How should this work complement / supplement other coordinated efforts?

  How to connect with university, foundation, and corporate approaches?
  

• What promising approaches + gaps do you see in the current approaches to supporting infrastructure like this?

Organizational advice

• What risks should we be looking out for at this stage of development?

  Fran: there are different success scenarios and business models for key stakeholders. For researchers, the success scenario has to do with being able to contribute to impact, professional development, etc. For the public sector, a success scenario is getting that info out there. The success scenarios for these 3 groups don’t align well. How do you actually distribute that?

• Who should we be talking to, as collaborators, advisors, sponsors?
  

Curation + knowledge-development advice

• What first-hand observations should be more accessible + reusable?

• What tools can augment human + scripted curation?

• What curatorial updates should existing communities share & synch?

• What models of the world should be developed more transparently?

• … (add your questions here)

General notes + transcripts

Cliff Lynch, CNI

• What infrastructure should academies legitimately be interested in?

• Analytical voodoo that’s not accessible to the academy

• Scholarly work bigger than the academy

• Case study: publication rights from faculty members

• Case study: faculty views on who owns research data — different than institutions

• Two missing audiences with different interests to represent: time and research productivity of readers and writers

  • career never moves on from first thing they publish is problematic

  • reader being able to assess knowledge with process rather than vetted, stable final product problematic

  • don’t understand when appropriate to use these and when not, and don’t know how to balance reader/author productivity and effectiveness

  • not necc. overlap — as author, telemetry good, as reader not

  • library wants data to make investment decisions

• Governance: you want some resilience from acquisition. Bulwarks, I say.

• Privacy: important part of academically controlled infra — reader privacy problem — who gets PubPub’s information?

• Active privacy, data retention, data fuzzing on ingestion: decide what the right thing is, learn how to do it.

• We’ve covered publishing + distribution; add preservation + maintenance

Peter Brantley, UC Davis

• Re-think scale and kinds of engagement we want to engage in

• How open do we want our software to be? Forkability, balance between benevolent dictator and “go use it if you can”

• What scale? Easy to over-build infrastructure. Too many dependencies or they tend to falter.

• Replicating complex systems like [full-breadth] cloud services: be very careful; choose appropriate pieces, like [services for preservation]. Easy to be overly ambitious and lead to frustration.

• Worth in the vertical, not just the horizontal — twitter could have been amazing if had just utilized existing IP protocols

• Working on UC integrated library system — idea that Davis could walk to Berkeley and be treated as part of UC — hard because built for campus, not consortium, nothing for sharing creds baked in. Counter-example: Eduroam, built on simple, old underlying tech — RADIUS and 802.1x

• Building researcher profile systems — highlight problems like inadequacies of datastores, points to simple but useful infra, ie, hard to know

  • To know that Fran has shifted roles and jobs, I need unique IDs for departments, specialities, organizations; every unit of [enterprise]. ORCID uses a proprietary one [Ringle?] OA alternative: ROAR

    • (aside: why did ORCID choose the former over ROAR?)

• What are the highly useful, easy to build, non-interdependent resources

  • Gabe Q: how do you balance this need with the “how open do we want software to be”

• No easy way to describe in metadata my activity in a dance — just the dance itself

  • Aside: trying to extend schema.org dramatically to cover little bits and pieces: new shapes, w/o creating many dependencies.

  • Aside (DH): wants to take advantage of these, not replace them. little bits are much more useful when interconnected.

Michele Kimpton, DPLA, Membership Consortia

• How do you attract stakeholders?

  • Governance to give input to roadmap, outreach, tech, etc.

  • DSpace: lots of forks and modules and uis and versions — resolved by regional centers of expertise

• How do you keep stakeholders aligned as it scales?

  • Splintering software leads to incompatibility, but you want openness

  • DSpace and Fedora merge — how do you differentiate from each other? How do I know you’re going to be around? Why aren’t you interoperable?

  • Different levels of stakeholders have different needs — ie provosts care about measuring contribution & promoting uni more than individual researchers

• How do projects scale outside of core adopters?

  • Think about this at the forefront before taking grant funding so you can continue when you’re done

  • Not based on size — let organizations decide what level they want and what benefits. Supporters —> Gold/Platinum

  • Allowed members to direct investment, resulted in siloed approach, little room for innovation — do we put a tax on member dollars so we have extra for innovation?

  • Governance was allowed at any dollar level, not strictly based on level

  • Projects had different boards than org as a whole — allows org to take more risks

  • On project boards, certain number of seats for each class, and election process

  • Industry: especially people contributing code. Could sit on steering groups but couldn’t have majority representation.

  • Kept in-kind different than financial revenue.

  • Board seats person-specific, not institution-specific

• How do you continue to innovative?

  • No R&D budget or tax on membership dollars

  • Institutions want to continue using what you’ve developed even as you move on

  • Can build R&D budgets outside of consortial fees or have a “tax” on membership — Lyrasis directed money from DB to “Catalyst” fund

• What is exit strategy?

  • Merge, partner, succession plan, etc.

• Understand where you are in the ecosystem

  • Lots in sub-$2mm space, hard to take off beyond immediate needs

• (SimplyE)

• Project Scale — DSpace/Fedora merger, saving overhead.

• “Data Engine” — an interlay taking data in and out…

• Asked everyone on the Board to write letters inviting people to write $10k checks; having a team asking is invaluable!

Open questions

• How does OSP think about sharing info w/o leading to power laws, overly strong follower effects, and most teachers using the same syllabus? A: they think about this a lot; some noising; some q’s to identify ‘cheaper than default’ choices, which adds to popular selections. [and how the options are visualized also affects choices/norms]

Infrastructure

• Travis: where are we trying to land, in this time of transition? Not just land, but paint a picture before so we have coherence, cultural instilling, norm development.

• Qs:

  • Where are the people/champions?

  • Build goals that you can measure — try to go through Heilmeier catechism — what does success look like, who will care, etc.

  • Article suggested by Robert Kiley: https://www.nature.com/articles/543179a

Breakout sessions + notes!

many taken here.

Argyle

• KFG can advocate for institutional consortia

• Partnership model — importance varies throughout institutions

• Dream team of folks in higher administration who, when they come together around a vision, can make things happen

• Value prop — has to be particular to needs of members — what value add do they bring to your campus and researchers (on the terms of your metrics of success) — ie endowment, US News & World Report Ranking — need to be able to frame their needs in terms of elevating those priorities

• How does KFG create the environment that it needs to be impactful? How do products fit together? What does a new roadmap look like over the next 5+ years.

• Short-term needs

  • clarification of everything. Goals, value, PubPub, Underlay, etc.

  • role for marketing these issues to much broader community — hard to make the case outside of ie library communities

  • just do advocacy — “stop with the white-paper shit”

• Not just about making sausage — who likes it and is it good for you?

• In favor of empire building. Be ambitious and pushy and claim the mantle. History of Clear, Scratch, DPLA, etc.

Chevron

• Highly apolitical, non-partisan agenda, operating at very high level

• Need a manifesto — declaration of the future of human knowledge, public and signed

• Travel grant — put on a roadshow, embed ourselves elsewhere in the academy, talking to provosts and admins at other institutions — not just the academy, but information producers outside of academics

• Advisory board — and quickly

• Target list of stakeholders — identifying the target list, especially hubs of target orgs that could make our visits or contacts more efficient

Mosaic

• Focus on low-level representations for knowledge that can be shared and avoid falling into tool blindness

• Distinguish access to info from access to meaningful knowledge

• Frame work as open knowledge production, facilitating innovation, helping people build more specific things on top

• Culture change — hugely important. A lot is collective action and politics. We should try some. Reach out to leaders of fields and try to help transform.

• Find early champions. Don’t just build. Ground in real use-cases from people doing good things and people who want to fund them.

Houndstooth

• Value of funders funding tools vs. things made on top of the tools

• Whether consortium based around tech stack we’re building or a lot of partners building related things

• Talking about researchers — need to show more examples

• Value prop for different players — would we be an acquirer of non-profits/startups in the space, providing them with infrastructure, appealing to be able to use KFG as a way to attract funding — don’t need to build surrounding aparatus. Part of what KFG would offer is diligence and more users.

• “Funders Day.” Non-profit media lab. People thinking of joining, funders, etc. could come and see what we’re working on. Universities could benefit from having a page showing their contributions.

• Mozilla model

• Importance of standards as a driver for policies

Takeaways

• Alexa: “open science by design.” By design = change how science is done. If you do that, no need to convince people to use a tool or share data after the fact. But if sharing knowledge and data is a by-product of how you do science, you’ve solved the problem and culture shift. IE e-notebooks — makes your work easier, and as a by-product, produces usable data.

• Greg: one of more important conversations. My passion is capturing diverse voices. Paradox: diversity vs. standards. How can we bring together scholars, presses, libraries together to advance the future of knowledge together? Super exciting. Historian — stress is breakdown of enlightenment tools and procedures — as we move forward, can accelerate the problems inherent in enlightenment or we can get it right. Huge benefits. But problems of colonialism. Inequality. Amplification at the expense of the marginalized. Future of knowledge, not libraries or presses or academia. Creating a big tent super important. Indigenous knowledge into the conversation important.

• Michael: on website, focus on standards is important. Hard to build workflows without policies/mandates for people to change behavior, and adoption of policies going to be a driven/market for the product you’re focusing on. Make sure there’s interest in that part. Critical for success. Having good sense of how quickly the space is changing. Not just from tech, but funding. Philanthropies shifting quickly towards sustainable models — and government. Moving away from funding projects that don’t have models built in. Critical for everything — how do you capture valuable tools nobody uses because it wasn’t marketed well, mandated, or integrated. How do we find value in those tools and incorporate them into a meaningful way?

• Amy: a lot of unexpected opportunities and challenges — but it resonated how important this is. In addition to specific recommendations about what shape that takes, inspiring and confidence-building for the group.