Navigation:
Documentation
Archive



Page Tree:

Child pages
  • W3- Consortial Model

This wiki space contains archival documentation of Project Bamboo, April 2008 - March 2013.

Skip to end of metadata
Go to start of metadata

Consortial Model

Table of Contents

Straw Consortial Model

Presentation slides
Video (mp4, 70 mb, 25 min)

  • Consortial model based on looking at projects that have gone on in open/community source
  • Based on worldwide conversations
  • Faculty and researchers
  • A lot of different ideas emerged
  • I tend to worry about success more than failure - it's a lot harder
  • If we were thinking a 3-year project, what do we do in year 4?
  • Within a few years, digital video has become part of everything we do
  • Lots of different communities are engaged in this
  • Bamboo is a project of multiple groups - lots of different disciplines, fields
  • Conversations between members of this community - these are new connections that might persist even if Bamboo goes away
  • Also looking at different technologies (saw a lot of that today - different approaches to problems)
  • We are a single community
  • Gray speckles: Bamboo community - people who have expressed interest in Bamboo (anyone who's subscribed to mailing list, or posted to wiki)
  • Lots of people who aren't here, and won't ever be in a meting like this
  • Includes every community we've talked about, and then some (heads of presses, lawyers)
  • Then there's yellow speckles: everyone in the room
  • Institutions who have put in time and resources - these are "members" - committing resources to better the community
  • Members actively contribute to success of Bamboo - but have different interests (some want to build software, others want to build community)
  • Way institutions manifest interest is different - invest programming/developer resources, other institutions (used to) say, "we have financial resources"; others say they have people to commit time to build community
  • Different dimension we can all bring to this project
  • Idea for the model comes from open source/crowdsourcing ideas inside/out of higher education
  • We need to learn from experiences so far in planning project
  • Have to develop model that can expand, contract, and evolve over time
  • Build on strengths of existing projects, create layers that can be contributed to and coordinated by members (not just people at the top)
  • Small pockets of activity contributing to larger whole
  • Leadership of this project comes from contributions of participants - Berkeley and Chicago aren't dictating, but taking synthesis of what we've heard/talked about and presenting it back
  • Leadership could come from community and be reinforced
  • Sustainability/reality of all this - we have to reduce individual institutional risk by distributing activity/services throughout community
  • No one institution has to bear burden for everyone, shared by multiple institutions in different ways
  • Problem of the Grand Repository - no one wants to run it, because everyone hates you when it breaks
  • "Core Partner" model - work with the community, community recommends what should be created/supported
  • Partners commit resources
  • Community comes together and evolves over time
  • Strengths:
    • Community provides input and guidance to those partners
    • Community free to come up with lots of ideas
    • Partners carry development burden
    • Track record of success in community source projects
    • Consensus is easier because smaller set of partners (can fight with 7 people instead of 700)
  • Weaknesses:
    • Prioritization can be a challenge ("why can't you do X?") - how do you decide what's important? > develop a mechanism for deciding/filtering priorities
    • Partners carry development burden > share with willing members of community
    • Community runs risk of disengaging from development process > "they don't care about me" if your idea doesn't happen, how do we establish multiple ways of participating in the project, take some ownership while contributing to the community as a whole
    • Lacks mechanisms for dealing with diversity > create mechanisms to promote and incorporate diversity (see Stories working group, bring new groups into conversation who might not be otherwise)
    • Basic model comes with community - practices in fields and disciplines
    • Success depends on engagement of members working to betterment of that community: Explore, Plan, Build - members participate in at least one
  • Explore: explores possibilities enabled or created by Bamboo
  • Figuring out what's needed - how people are using the technology
  • W1 and W2 are parts of "explore"
  • Focus on outreach, education, community support, scholarly networking
  • Developing, growing, nurturing Bamboo community - adjusting as necessary
  • Needs to be a filtering process: Plan layer
    • Planning infrastructure and standards (loosely defined, best practices, guidelines, how we need to interact to share data/applications/ways we work)
  • Takes input from Explore and prioritizes those activities
  • Figures out what needs to be done and expresses what should be part of the community and portfolio of things Bamboo members use
  • Filters - "have you heard of...?"
  • "Bamboo should come up with a way of doing text encoding!" Explore says "Yes!" Plan says "Maybe we should engage a conversation with TEI"
  • Builds connections with communities, that didn't exist before
  • Leverage strengths of that community into Bamboo
  • Liaison role - formally intersects with "plan" layer
  • Formal conversations, rather than informal conversations that may not get too far
  • Smaller set of dots in "Plan" layer
  • Build - actually builds stuff, takes prioritized list and builds that, and makes it available (thinks a lot about technology layer that will have to run underneath Bamboo)
  • Need to have some kind of leadership in there-- if you look at stacked dots, this institution may say "I'll step forward to take leadership role in coordinating activity in a particular layer (build, or plan)"
  • Then we pick up back on the partner model - at some level you have to be able to make decisions for Bamboo
  • Lead institutions would invest resources to wrangle, coordinate and guide activities within a layer
  • Lead institutions should reflect activities going on in that layer
  • Lots of things that could potentially be going on - those things have to be represented by leadership that's there
  • Lead institutions become "Bamboo Partners" - burden of moving PB as a whole forward, become part of core set - these become leadership team and move tings forward
  • Model is based on members working towards betterment of broader community
  • 3 layers
  • This is a circle- everything feeds back into the community for assessment of value
  • "Members" shape and undertake individual aspects of leadership at a small scale (for a service, or something for leading professional development activities for scholars)
  • Members take/undertake that at each layer
  • As members take more responsibility, these become partners
  • Partners guide and shape the direction of Bamboo's members

Q & A

Q: In build layer, not just build but also "bamboozling", buying, packing, installing?

  • Yes
    Q: talking, marketing, explaining, educating is important - do we need an outreach layer?
  • Envisioned in 3 places-- can happen at explore layer, at plan layer - formal expression of what Bamboo chooses to do (to head a Bamboo service), leadership layer - important for all projects, who can represent BAMBOO? We all do, but at some point someone wants to point to one or more institutions that can speak on behalf of it

Q: 5 sides of activities, but 3 core things Bamboo stands for
I don't see stories/demonstrator labs as distinct from what would feed into the first three points
7-10 years... maybe longer?

  • Reasonable to look at combining/looking at it differently
  • Pulling stories piece and labs piece out differently (they could be combined) was based on trying to think about what we'd heard and how to build a more compelling value proposition for institutions
  • This sets us up for tomorrow's discussion, too
  • Are we organizing/conceptualizing this right? It's important to come to a reasonable way to do that, and move forward

Q: You mentioned examples of partner model - names?

  • Kuali - lots of good ideas in running a project like this
  • Sakai - the other I'm most familiar with
  • Each has had their strengths; many years later, observing initial work of Sakai, had a lot of voices coming at it - had to sort through those things and make sense
  • Some areas it worked, some areas it didn't
  • Different animal in Kuali - community itself is a different community (lots of administrators, thinking about business processes, wired more like IT folks)
  • Much more known outcome at the end of it - we're enabling connections/capacity for innovation, different than trying to develop next-generation student system
  • How do we embrace what's there and has already been done, then think about diversity?
  • If anyone has community-source activity/tech activities or other consortial models that are wrestling with diversity vs. development issue, would like to hear about it
  • Greg Jackson: Expanding on what Chad was saying, this touches on something subtle and difficult
    • Two related points
    • Kuali story: continuing attempt between Kuali and University of Chicago comptroller to have conversation - "Here's things we need, and when we need them - can we get it?" Kuali: "Well, this is a collaboration - what will come is product of collaboration. Yes, it'll happen if you persuade everyone." "I can't get into it unless I know." "Unless you're in, we can't guarantee anything."
    • Different than vendor/customer relationships - real dilemma; there'll not be a satisfactory resolution in a way
    • Lots of linguistic issues floating around: different between "plan" and "promise"
    • Plan - best thought about best steps to get to a given goal
    • Most of us don't do what exactly we planned, sometimes to a different goal
    • Can't confuse plan with promises to achieve particular ends, and proceed in particular ways
    • Nature of implementation plan/proposal - have to remember both ambiguities

Q: Natural to say that leadership comes from partners who are very active on multiple levels, but there's a tension- you don't necessarily end up w/ a result used by anyone else
Our provost isn't into planning/building/exploring the IT structure we use, but if we try to bypass the provost's leadership, we can't get anything done
Flexible definition of "member" - hoping we don't limit to institutions (could have learned societies, other groups that won't necessarily be involved in all phases but will be treated as members and have some leadership)

  • CJK: Yes - we started looking at industry partners, etc. - we'll hear from Microsoft tomorrow about their views of us
  • Not everyone has to be a formal member to contribute to Bamboo
  • Liaisons and affiliations - formalizing connections that already permeate this project
  • Other opportunities to engage, too
  • Consortium of institutions coming together; CIC could become a member
  • No rigid definition now- we define what those definitions are

Q: Relationship between consortium and implementation phase - will consortium be enough?
Will you set up a consortium between participating institutions, then use as a springboard for various applications for funding?
This could give more independence/flexibility to consortium; implications for ownership of Bamboo ground

  • DAG: Discussions about writing this proposal so it's a framing proposal which allows different kinds of work to happen, different proposals to interact
    • We have to clarify that here, and with the foundation
  • Chris Mackie: We have no ambitions to own the brand Bamboo
    • I can't see our funding being sufficient to sustain this
    • We hope to contribute catalytic resources, but we've assumed all along that additional build-outs (connecting of Bamboo to other connections) would be funded by a mix of places
    • Anywhere anyone can find money
    • No need to worry about our engagement - we'll be evaluating on our best judgment for sustainability, plausibility given resources, and our board will evaluate based on whether foci line up with the Arts and Humanities scholarship they're inclined to support
  • CJK: I'd measure success if individual projects could seek funding knowing there's resources and capabilities and community behind those projects to make them successful/help them be successful
    • Bamboo helps give those projects initial leverage
    • There's not One Grand Funding Moment, but creating opportunities for a lot of innovation to occur

Q: You might've left this vague for tomorrow, but there's a lot of relationships between institutions, prospect of co-dependency
Have to have a view of trustworthiness of other institutions
Where are you on spectrum of formalizing relationships - are you thinking legal contracts? Process of due diligence?

  • We talked about that a long time ago
  • We have to deal with the "what's in it for me" problem, for all the communities
  • If we want to build something that's sustainable, you need to look at/understand/sort out a way where if I choose to use your service, you know it'll be there
  • Imagine our network problems happening in a major conference or classroom
  • You want it to be an ISP outage, not due to a service you've counted on having a hiccup and shutting down
  • We haven't figured out how to do that yet
  • There are certain services you're going to depend on - you know it'll be there because there's copies around the world, data's always there
  • But I'm a researcher w/ a cool thing under my desk
  • If another scholar or service provider connects to that, I need to know about that service -it may be the best, but if it's R&D I need to know (not discover) the risks
  • How do we establish that kind of "contract" so folks in the community know what you're getting into

Q: The way you describe the consortial model, it puts a lot of authority into that group who are probably builders and central and participating in the layers
Can we build in checks and balances by having a council or board or something that comes out of that membership that drives all of this - also includes diversity (elections, turnover)
Harder to change a build group than a board

  • CJK - I will be going more into that; I have one notion of advisory (which I'm now rethinking) - this may be different than what I present in two days
  • Issue of checks and balances and fairness - always a risk of one institution dominating all
  • There's a tendency to say "build layer here is the most important" - there are 3 pieces, all of which are incredibly important
  • I view each as just as important as all the others

Details of Straw Consortial Model

Presentation slides
Video (mp4, 73 mb, 26 min)

  • Members leading activity for the broader community
  • Three layers, explore-plan-build
  • Members undertake activities in each layer-- similar to working groups
  • Members take on responsibilities and leadership and potentially become partners
  • Partners guide and shape membership of Bamboo
  • "How does this work if everyone's a flat layer, where's the coordination?"
  • Exploring the possibilities created by Bamboo, trying to understand what community is doing, how things are working, what changes are happening
  • Outreach and support reaching out to extended community
  • Developing/growing community over time
  • Int support working group- fits nicely into this space
  • Let's say there's a group interested in working on different aspects of scholarly networking-- that could be a cluster in one dot of institutions
  • Cluster of institutions to work on a topic-- either get funding or maybe have funding
  • Problem: how does that relate to Bamboo? Need some sort of coordination activity
  • Working group chairs could also be institutions-- coordinate clusters of activity
  • One cluster could be around scholarly networking, community development
  • Working with interesting text analysis tool-- this leads to affinity group around that, then says "you need to come up with a new text encoding scheme"
  • You want to have a check and balance in the process-- plan functions as a role in that, ta filter between build and explore -- is there a society? or an organization with a schema?
    • Builds connections, not reinventing the wheel
  • Only after you've gone through that process, realize if you need to make something
  • Prioritizes activities for Bamboo-- what should we be investing in
  • Expression of standards guidelines and practices in plan layer00 not defining and writing a standard
  • Standards and practices that are formalized-- if you're trying to start a project, recommend using TEI
  • Could put together a profile for particular fields in TEI, encode in this particular way
  • Guidelines and practices: same sort of idea, but ways that will come from the community about how to approach and tackle problems
  • Evolving methodologies, tools-- how do we get those out to the community?
  • Intersection w/ a lot of other humanities activity around the world
  • CenterNet-- what are your best practices? Can we link to/inform community about those?
  • Notion of liaisons: formal definition of liaison with Bamboo
  • Lots of us have "pub conversations" - not necessarily = formal discussion and decision making about a process or problem
  • In plan layer, opportunity for institutions or community members to be a forma lliaison between organizations
  • Bamboo and TEI, for example
  • If there's a recommendation that come out that says there should be an extension or profile of TEI spec that should be used, there can be something formally taken to TEI
  • Very important formal relationship for being in this community
  • Acts as a filter -- not building things you don't necessarily need
  • Some great ideas lead right into a brick wall - the last thing we need is a lot of brick walls
  • Standards guidelines & best practices, and architecture & services
  • Services- what architecture is needed, what share services are needed
  • Standards- how do we promote interoperability through adoption of standards
  • Overlap between them, but referring to them as different clusters
  • Two institutions may wrangle those particular spheres of the plan layer and provide leadership
  • Build makes the stuff-- takes what's already written and deploys for community too
  • Build layer is responsible for that tech infrastructure layer proposed as part of PB
  • Build layer is responsible for thinking of architecture and infrastructure under the clouds within clouds
  • Executes priorities established by "Plan"
  • Not as many institutions interested in that responsibility-- they work together to drive build layer forward
  • Three layers that depend on one another
  • Whatever build is done will affect the community, which generates new practice,s which leads to new recommendations - leads to a cycle as community/Bamoo evolve
  • Liaisons
  • Formal liaisons is critical-- there's existing consortia, but there's also been discussion about how areas outside the US
  • The way a lot of projects happen outside the US - driven by nationa intitiative s and large regional activities
  • How does PB engage with those? - regional networks
  • Some people have multiple hats as members of organizations; "for the sake of Bamboo, I'm with this organization, but I may sometimes represent institution if necessary"
  • Can participate in PB in a focused fashion
  • Liaisons interact formally w/ plan layer; engaged in other places (we all participate and contribute in different places) - formal connection comes in with plan layer
  • Long list of potential liaison organizations
  • Formal liaisons interact between memberships, fits into fabric of plan
  • Strong connections of different layers of different lead institutions
  • How idea of partners comes in - 3-6 (at least 3 because of 3 layers)
  • Sit on some sort of leadership thing ("the Board")
  • Each partner represents the layer it's leading, reflecting what's coming from that layer
  • Big Bamboo Direction, what we should be doing as a large organization
  • Need a Bamboo "Office" - there's a lot to running this, it's a lot to manage/maintain
  • Institution that administers the business side, deals with large community activity coordination work, but doesn't lead any of the layers
  • An institution not buried by leading something else - a balance across the different layers
  • Fills role of neutral party, chairs the "Board", acts as the "executive director"
  • Bamboo "Board" - works together to ensure PB is headed in the right direction, however "right" is defined - should be clear if it's coming out of layers of community
  • Leads Bamboo-wide activities and initiatives
  • You don't have to go all the way through the chain to make something little on the edges - flexibility for things to happen, and as things bubble up, if things become more important we can formalize it
  • How do we get activities out of the system?
  • "We don't need to be doing this anymore, the world has changed" - passing things off, deprecating it, etc
  • Defining PB strategy - what does it do in the long term (beyond the next 6 months; 2-4-6 years)
  • Responsible for PB and strives to do what's "right" for the community
  • More or less going in a direction we more or less agree on
  • Not going to be 100% agreement on everything; if we can achieve general consensus, we should be successful; interesting things can happen at the edge
  • Each piece is a different layer that works together as one whole
  • In reality, it's bound to be a lot simpler than that, but we need to sort through as part of the consortium what is the right balance at this stage of the project for investment in resources and leadership in build/plan/explore layers
  • Adjust accordingly
  • Over time, taking input from the community, this can grow or contract at any layer size
  • Right now, we may decide there's a lot we ned to do from a tech development standpoint, but in a few years we may have infrastructure inp lace, and resources hsould go to exploring and planning
  • Continue to adjust over time-- decoupling administratively distinct activities so they can grow and change
  • Different people/kinds of organizations interact differently on those layers
  • Administrative model, not necessarily cultural model
  • Region-based model for a limited time?
  • Important that it connects back in to a way we can manage this thing-- a lot of coordination and wrangling that needs to go on

Q & A

Q: You said that priorities from build comes from other layer, and there's institutions already set up to do it
Will build institutions be influenced by their own need?

  • CJK: You have to have the plan layer before the build layer-- can't build w/o a plan
  • Maybe I've been involved in too many construction projects
  • There could be a set of institutions providing certain development resources on build layer, nearly a project management coordination role that's leadership of that layer, but based on each aspect that comes out of plan, you need to launch projects based on that
  • Works with "build" to find, identify resources appropriate across build community to make it happen
  • Not "one institution does one thing"
  • Q: So, do plan layer have to do work by Sept?** CJK: We're rolling all this together right now
    • DAG: We're laying out a high-level outline of the proposal right now
    • Goal is that we're going to have more detail about what PB will do in a 3-year timeframe
    • This will inform what consortial model looks like
    • Putting out a fair amount of abstraction right now; that can lead to confusion but we're trying to tease out a set of work that can sustain that over the first three years
    • In proposal - enough institutional information early enough so institutions can say "this is the role Iwant to commit to"

Q: Two potential sources of confusion: seems like there's a kind of congruence between community/membe/partners and on the other explore/plan/build
Plan layer is where real decision making is going to happen?
Already defeats community-members-partners pyramid
Since temporal structure of planning project has been chronological explore-plan-build

  • CJK: How can we continue to evolve and move forward?
  • Innovation is when people have an idea and want to act on it quickly
  • If we really believe our core value proposition about enhancing arts and humanities, we have to allow that to happen
  • Can't be tied closely with every major administrative move
  • Some aspects have to operate and move in that lockstep movement
  • How do you separate the two so you can have activities that move quickly,
  • If things become important for community, they can be moved into and formalized as part of what PB is
  • Sustainability part of things
  • If everyone starts depending on ARTFL's sequencer service, as we think about sharing resources (we saw this on Monday - when one piece of infrastructure breaks)
  • It has to be there when you need it - need to know in advance if it's not
    if you're comfortable running your thing knowing it may go away tomorrow, but if projects are depending on it, that may necessarily need  a formality

Q: re John's question
Whether an inst might not want to commit resources necessary to build something unless it had a hand in planning what it was helping to build - is the answer that a single institution participate as both build/plan
Also leaves room for institution stepping up after plan has defined something

  • CJK: Yes
  • Explore - humanities, plan - IT, build, Comp Sci

Q: This is a straw model - how do people feel about separating these things
Strength of the line in the diagram = amount of resources
Will this structure work? Then we sort through the resource issue
In our discussion, folks at tables can bring up those issues
Is this even workable? If this is workable, then we can start talking about commitment layers

Q: Exploration labs
At the end of the process in 'utilization', where rubber hits the road, community learns how to use Bamboo
Since exploration lab could be in explore part, too
 Could you talk more about the exploration lab and its role? This would probably be an iterative process

  • DAG: Exploration lab at the beginning, middle and end
  • Model valuable to many local institutions
  • A place where all this can happen - for generating ideas that might push forward planning/building
  • Looking at how that works with tools and services
  • Share stories there, train people in tools
  • How can this be manifested in institutions?
  • Explore commitment- building more connections between learning and hsaring and having info go both ways
  • CJK: Conceivably, an exploration lab could exist in the diagram where lines intersect

Q: We are fully consortial institution-- faculties drawn from various institutions
We had to develop complicated consortial model
Codify that consortial model extremely carefully - a group of people in this room are the "original gangsters" - they'll be gone someday, and people lose sense of connection to what started this
People come in at the community layer level and not be clear about partners/leadership team
Need a doc we can all use that's fairly clear
Office/exec director model - if it's a wedding cake, it's a fairly traditional couple sitting on top
Is that the right model for the kind of broad-based community that sits at the broader level?
This might lead to some alienation

  • CJK: Help us define a non-traditional couple

Q: Don't know how resources would be distributed in an ideal world
  Building is going to cost more than planning/exploring, so this would mean more resources asked for/going to people for that

  • CJK: Depends on resources-- money is one resources, but so is time
  • What are the commitment levels at each place, defining resources needed for those
  • Very little dollar from the project
    Q: Won't leadership tend t skew towards building because of financial skew?
    To evaluate this, need to do it without those considerations
  • CJK: by decomposing into different layers, trying to set up checks and balances
  • I'm also worried this could be building a lot of technology, and it decouples with community
  • Berkeley workshop, 1a - "the normalization of research makes it irrelevant"
  • We need your input; there may be another model altogether of structuring this
  • Important to maintain balance, and don't have handful of institutions dominating because of having more resources
  • There's a reality that in order to make tech, will take serious resource commitment but that shouldn't be true for every aspect/movement of PB

Q: Notion of "build" as inclusive; now we're talking about building tools?

  • CJK: Build focused on infrastructures-- creating ways in which we can share services among community
  • Tool development ends up happening largely out on the periphery and among people who are part of the community
  • Groups of people on other ends of the periphery should be able to use everything the community is providing
  • "Cloud" - still very new, how do we share/expose those resources
  • Not necessarily "how am I going to make a tool for text analysis" - that's not the place of "build"
    Q: Facilitating networks/professional networks
    There are other deliverables
  • DAG: Scholarly network is important, overlap with the 'build' side
    clear description of work and what parts need to be planned/built (incl services, scholarly networking, stories, exploratory labs), clear description of resource requirements, clear description of what the different roles are that are possible in the consortium to do that
  • We've given 2 pieces in fuzzy form with very little information about resource piece
  • Hard to see in relationships
  • Using discussion from this workshop to flesh all three of those out and present in more detail going forward to W4
  • Also there's build requirements around scholarly networking, sustaining relationships within/between institutions
  • Next step of putting together verbs and end-results that go with those verbs

Q: Appreciate efforts for creating hierarchy for action and ecology for things to work in
Where do you see funding agencies in this ecology, are there other pieces we need to think about but aren't here and need to be modeled

  • CJK: Funding agencies are hard in terms of fitting them, but they do fit
  • Do you have a formal liaison w/ a funding agency; a little weird, maybe?
  • As we walk through this, will be important for everyone to identify 'will this work for a particular community' - University press community comes up more and more
  • Is the role in which the univ press engages a separate member entity or a separate institution
  • Is it plan layer with formal liaison relationship?
  • Lots of things that all of us as we flesh out the details have to talk with and wrestle with
  • We won't get all those dimensions in right away
  • Make it easy for those connections to be made w/o turning entire model upside down
  • If you look at cross-section of institutions-- industry can engage
  • Industry could be a "yellow dot", just like the fine arts museums of San Francisco
  • Engaging from a funding standpoint, here's things that'd be possible for the community to do
  • Not really here, it's not a member, but affiliation is odd because what are you affiliating with? - but you have to have relationships

Reflection: Greg Jackson - Chief Information Officer, University of Chicago

Presentation slides (pdf)
Video (mp4, 35 mb, 13 min)

  • Where are we? - there's a period where people expand the choice set, and makes it more and more complicated where it seems impossible
  • W3 is the middle of 5: this is the point where we move toward choices as opposed to expanding choices
  • At a point where we can't just keep coming up with pretexts why it's not time to go inside
  • 26 years, 3 1/2 careers ago: National Institute of Education putting out a competition for a research center for how to use computers in schools
    • Harvard: trying to figure out how to get dean to spring for money for it
    • Dean said "why don't you compete for the federal grant?"
    • Call for proposals didn't have a good take on substantial issues, but said there should be a specific regional/national approach to issues
    • Need to work on consortium question; we did something similar to Chad's complicated diagram that depended on a lot of different institutions
    • But it became clear that as you look at it, if you look at all those members of consortium, resources likely to be available, would be tough choices
  • Two things:
    • First seemed fatal - consortium fell apart, split into 3 consortia, wanted to bid against one another (we didn't want to get into a political thing, we just wanted a little money from the dean, and we don't have any particular credentials)
      Told dean it wouldn't work, but dean told us to bid on it
    • Sat down and said "if we have to bid, instead of looking at what NIE wants, let's figure out what the right way to think about it is" - talking with people, wrote a proposal that was unresponsive to RFP and sent it in, and a month and a half later we won
  • Lesson learned: important to think clearly and crystallize thoughts and get them forward
  • As we look to the next few months, this is what we have to do-- make sure everyone in the enterprise has a piece of the thinking and resources
  • Look for the strong ideas
  • 3 kinds of activity being talked about-- these have different attributes re: resource calls
  • Choosing among different kinds of activities call for different kinds of resources - don't choose things that eat all the resources so things get eaten alive
  1. the kind of thing going on here, immensely valuable even though it's not building tools: "the forum"
    • Convening folks who are engaged in these issues, good thinkers, to work together and mutually support and move forward
  2. "grid" - not technical notion, but idea of standards and notions; what we do together is not to build things, but to create ecology, frame it so people know what their pieces have to fit in
    • Grids can be rigid and laid out, but this is Dutch mathematician's attempt to explain Escher's "Picture Gallery"
    • May have singularities, problems that can't be solved, don't fit - but no excuse for not doing it
  3. "factory" - literal sense of building
    • Getting together, hiring programmers, making things
    • These three kinds of activities call for different kinds of resource sharing between "center" (Bamboo itself) and members/partners/other pieces.
    • If we think about forum: about $50k of central funds to do a workshop
    • Costs each person's institutions $1000 for people to come here, $120k spent by institutions to have this forum
    • Things like this tend to lead to a balance where 3/4 comes from institutions, 1/4 comes from the center
    • Central contribution is fraction of the whole
  • Grid: something that's more balanced - integrative thinkers, standards setters, in addition to ordinary logistics
  • Call for more center resources
  • Factory: central piece becomes quite a lot larger
  • Why is this important? As we think about what's beyond the planning phase, we have to think how much resource to put into these kinds of activities
  • We have to be thinking in terms of finite resources - there's some amount there is, and no amount of rhetoric changes that; changing how it's allocated
  • If we decide factory is an important way to go, and total Bamboo resources is fixed, we're deciding we don't have much resources to do the other
  • Making these bigger or smaller changes what can be done centrally; detailed trade-off
  • Size doesn't depend on what we say, its externalities
  • How we allocate resources w/in that is important
  • Any time we're making arguments we have to understand we're making arguments against funding the others
  • Tough set of choices; we keep talking/announcing principles because these are going to be tough questions that will be hard to face
  • "It's dark enough. Let's go!"
  • 4 and 5- we have to make these choices, then make them the best ideas we can, for clear thinking about problems

Q&A

Q: Define grid more clearly?

  • Standards and requirements, and API - but this understates it
  • There's choices people can make, and we have to guide those choices
  • Someone's got to do that
  • This is where advantage in Hum Tech is: still have opportunity to do that
  • Science computing: even though there's resources, it's hugely inefficient because everyone did their own thing instead of doing it a priori

Q: Can we map store, plan and build onto forum, factory... etc

  • explore > forum
  • build > factory
    Q: where in the nomenclature does common direction setting come in?
  • It maps squishiness in the middle

Q: We already know certain things about the grid (electrical, water, etc) is that part of the grid how we're looking at it?

  • When I'm talking about grid I'm not thinking hardware/software infrastructure
    Q: There's certain knowns already - we're using computers; question of normalization of humanities
  • The grid we already have in place that has nothing to do with Bamboo will dictate some of what can happen
  • I mean something above knowing how computers work; if you turn the clock back to how one would represent images on the web - early on, that wasn't a well-solved problem (bitmaps dependent on particular technology); jpeg = joint photographic experts group
  • 'What should we worry about' - quality vs size
  • Set out standards for how to think about that
  • For a long time, that let people handle images so they could say "problem solved"
  • That's the kind of grid I'm talking about
  • There are other ways one could do it, but they made a reasonable choice, everyone followed it, and it saved a lot of time

Q: I like the model, it starts to get us involved with concerns and tradeoffs
It separates out into discrete resource streams-- what I've been hearing, is how tied one is with the others
The model has to be a lot more complicated before we really understand where resources are going and what tradeoffs are

  • Yes, of course tradeoffs are more subtle
  • I keep trying to draw charts in PowerPoint, but it doesn't represent my organization-- melting ice cream in five flavors
  • No good graphical way even with transparency and circles to get it to work
  • Not resolvable within our lifetimes - we have to make a gut sense of what the right answer is even though we don't know and proceed
  • Doomed to failure if we have to complete the model before we proceed
  • Professional judgment - now the answer even if you can't prove it

Q: When you make a determination about commitment of resource, the impact on vying from potential community members
Worth pointing out - people want to get the biggest pile of money, but this could be at the cost of the community

  • We want enough info going into W4 so they can talk w/ institutions with what they're interested in

Q: This framework signals to me that there might be design implications of how we structure this
If resource base is forum oriented more than factory oriented, how we approach building may be more decentralized

  • May also help us think about who we approach for what resources?
    There may be something beyond institution and Bamboo with resources for only certain things

Table Discussions of Consortial Model, Part 2

Tables were asked to address the following questions in their discussion: Do I see a place for my institution or organization? Are there critical gaps that should be addressed? Can the model work?

Table 1

  • Matching individual organization/org timelines to timelines that come out of consortial plan/build/explore
  • One of the issues of a local inst participating is they might have grant funding for working on some aspect of a project, be nice to part of a consortium to build it
  • That kind of mismatch of local timelines, how local groups would be able to match their timelines with their projects up to larger group being able to do it is a question
  • Also, aligning priorities-- suggestion that whatever inst might think it's gonna do, proportion to how we'll they can show their local priorities can be aligned with larger priorities
  • Discussion about exploratory labs - saw institutions being ones that might primarily come in as exploratory labs
  • Couple people thought "have to be careful with lab" - we didn't take it to mean administrative structure, space, machines, etc - more like exploratory projects
  • There are institutions that come out a little bit with some form of outreach
  • Still some general confusion about what might be meant by that
  • If lab does something, how is it going to disseminate out?
  • Planning layer needs to be done so in some sense it's open to whole community
  • There are places where people might not be able to allocate resources to build something but are interested
  • Tried to poll institutions about what different layers they might participate in, 3 said they could see someone doing any of them; some mentioned the liaison role

Table 2

  • Lack of business model
  • Questions about contributions of resources
  • Gaps we talked about
  • Poll aroud faculty and how they feel about where we're at -
  • Faculty not feeling comforatble w/ language we're using as focus of work
  • "So IT oriented, not sure of concrete value of what they're going to get out of this"
  • How will a scholar/teacher see how to use all this
  • Value propositions - not clear how institutions could see moving ahead
  • We oculd see our institutions participating, even smaller institutions
    about faculty pushback - the faculty here feel comfortable, but pitching to other faculty is a problem

Table 3

  • Liberal arts colleges
  • Governance pyramid may enforce hierarchy we see among higher ed institutions
  • Governance structure should have agility necessary for innovation
  • How does gov structure show value of diversity - specify representation?
  • Didn't see a process for generating assessment of organization, adjustment based on assessment
  • Haven't seen a lot of group votes, e-mail votes, methods for everyone to provide feedback other than e-mails and wiki
  • May want to reach out about that
  • Comment of the last table: faculty are the unconnected dots
  • Bamboo should connect the dots
  • We need to make cons model support and enforce this goal
  • Should there be a clear communication goal specified in the plan
  • A lot of faculty needs would be/are getting solved in variety of ways
  • We all need bamboo-like conversations at local institutions
  • PB has to make sure it adds value to these conversations; if not, members need to effect change

Table 4

  • Would my institution join? Business case to take to the dean
  • Cross-cutting; you have a 4-layered cake but things have to cross
  • Evangelist; t-shirts, swag, things that get people to come in and stick to it rather than "one more thing we have to know about and do"
  • Will consortium work? Planning layer is a concern - if it's small, you have a bottleneck between cool community stuff and builders
  • Agile - need to move planning one down, so what planners do is get builders very close in sticking to labs and stuff like that, rapid and iterative development

Table 5

  • Interesting question particularly about the question of who the community is
  • Is it the people in the room or the institutions we represent?
  • Tension between that and that goes back to question of individuals as members
  • In our discussion, had a long talk about relationship this has to hum faculty and how language would not be a good sales pitch to humanities faculty and building a kind of force on the campuses to go to the provost
  • "We really need this"
  • Other issues, some of which have already come up
  • One thing that would help clarify: ultimate goal of advancing hum and arts research through new tech has appeal to humanities faculty but goal of PB, deliverable, are services which are great and networking, all is valuable in preparing the way for ultimate goal
  • There's a gap between goal of what PB can contribute to ultimate goal, and other pieces that will make the ultimate goal happen may not be in PB's provenance
  • PB can connect to, support those things - help explain to hum faculty how this is valuable, even fi it itself isn't going to fund dig research labs
  • Those labs are really the implementation where everything comes to fruit in this project
  • Not explore in sense of needs assessment; don't want to call it exploratory labs, confuses the matter

Table 6

  • I-school faculty, small table
  • Institution that's scholar-driven
  • Institution has a place - we see ourselves as perfect spot for citing the exploratory labs (we have cheap, good labor) we are well connected to other departments on campus where people may come from to use those services
  • Critical gaps that could be addressed: since funding model is very unclear now, and sounds like the farther you are into gray area the less likely that you'll be getting any funding from PB, but maybe using PB services, maybe a mechanism for some sort of cost-shared role for PB to say "we're providing this service, ti's worth X, you can fold into your other grants to use as a support for your related work"
  • Can model work? Probably, but it's going to change constantly and dynaically as it moves forward

Table 7

  • Started out with feeling that we're confounding project and consortium, it's all intertwined
  • Important to tease them out to be more articulate about needs
  • Important to have assessment mechanisms in each layer to know we're on track
  • Diversity in terms of institutional types and other forms of diversity
  • Important to identify successful exemplars; scalability, how we can do that in A&H
  • Resource issue: if we have a fixed pot of money, how to allocate between grid and other forms?
  • Do we emphasize importance of getting people together, building tools?

Table 8

  • Attempted to answer questions, but conversation started in reverse
  • Can model work?
  • No clear answers
  • One thing that was similar to what was said - we're muddling up two different issues
  • Project phase and project, community, consortium, what's beyond the project phase
  • What comes after, continuum is more important?
  • We're on the cusp of having to make decision if we're a commune or constitutional assembly - one requiring more resource to sustain and manage than the other
  • Discussion of models of subscription, how it can be supported from w/in community rather than cenral project funding
  • Getting voluntary collaboration between institutions, gaining trust is hard
  • Do we see a place for our institutions? Everyone said yes, but wasn't sure which place

Table 9

  • Small, diverse table re: institutions
  • Critical gaps - saw unclarity of value proposition as critical gap
  • What are we going to put in/get out
  • Discussion earlier - obvious ways in which we each contribute (resources: money, people time = money, expertise = people time = money) might get back tools/services, but tools/services are nothing w/o content
  • In selling the plan, we need to remember we are all providers of content and collections
  • Content is being neglected in the big picture
  • Tools/service facilitate access to everyone's content - implies being plugged into standards PB might define; leads to implication that whatever else it is, PB grid is a grid of rights of some kind; sharing content rights
  • Content becomes another currency of involvement and participation
  • This could be a hook for individuals and institutions alike at every scale

Table 10

  • A lot of "not yet" answers - hinged on value proposition/business model
  • Hard time engaging our own faculty with elevator conversation about what PB will deliver
  • Part of a lack of clear agreement that we are producing tools to build infrastructure vs. supplying infrastructure, supplying services, supplying cloud - no firm yet understanding which it is
  • How those resources are valued depends on the outcome
  • Relatedly, depending on what level PB is proposing to deliver, may be other resources in play from these institutions in a support role, not in the diagram everywhere
  • If there is a cloud, how is that valued?
  • If the scholarly network tool is critical to PB and its' hosted somewhere, how is that valued?
  • Concern about early vs late adopters - some institutions which can't contribute now but could contribute during build phase; how do you value input of early/late adopters, balancing incentives of helping out in the beginning vs need to grow a community

Table 11

  • Introduce element of time - one of concerns about cons model, it's very elaborate, centralization about it; given discussions of limited resources, it seems there's implicit assumption that there's a limited timeframe in which PB has to get something done
  • This is backwards - we'll end up with a more simplified model in reality
  • Opening discussion of this kind of model will run risk of calcifying view of what PB should be rather than allowing qualities to emerge more organically
  • If we set our sights within limits of resources, focus on small, clearly defined goals to begin with, this is how you earn right to build towards that kind of model
  • We build up to that model
  • Also solves question of how to sell to institutions - we've achieved these four small things rather than tackling a large thing
  • Tempting to go for the home run, but without undercutting our own potential, doing something achievable in the beginning will lead up to greater success

Other questions

  • Q: Is there a place for for Berkeley and Chicago?
  • CJK: Chicago doe snot have capacity to be actively engaged in a build project
    • Difficult for us to be big open source software developers
    • Planning and explore conceptually is our strength
    • Other opportunities- that's where I envision Chicago
  • DAG: Moving towards wrapping up our discussion
    • We wouldn't be putting this kind of investment in stewarding forward the planning if we didn't think it was highly valuable for us
    • What's the best place for us? premature on that
    • Commitment is very strong across the board at Berkeley, from A&H and technology, library, information science, computer science
    • Opportunity is right to bring these communities together at our campus and elsewhere
    • Committed to making larger vision work; exact roles take more time
  • Articulation of value propositions is important, understanding of what it means for people/institutions to be involved
  • Need arguments to take back to their institutions, make the case with people there
  • We recognize that in working group; want to rename as "strategic communications" - arguments for involvement, roadmap for involvement and contribution - what are the steps an institution might follow as they participate in PB
  • Practical point about the timescale - we go into the work plan of implementation project, we need to do that NOW, before the proposal
  • Can be part of the work plan to refine, improve, etc but we need a statement ASAP
  • DAG: we've heard that from many institutions, feeling the same need
    want to have a little time in the net week or two to digest
    move towards having first draft before W4 circulated out so inst can have that and begin those conversations
  • Use W4, refine it, then another iteration between W4 and W5
  • Call from community input
  • Call for wiki-based value propositions
  • No labels