VERSO Playbook

VERSO OSPO Playbook

Creating an Academic OSPO is a complicated and compelling process. This is going to be an ever growing collection of resources created by VERSO to tell with upmost transparency to help facilitate others trying to create Open Source Program Offices at their location. If you have questions or want to learn more, please reach out through the contact form.

Summary

Created in 2022, we approached the creation of the OSPO as a design problem. This meant that we started by building a community map and creating a set of interview questions that we asked of the staff and faculty. Of 150+ people emails we were able to get 7 interviews (at the time school was not in session). This identified several themes that we would have to address:

 

  1. Lack of familiarity with Open Work, Open Source and related terms
  2. Not identifying as part of Open Work or even as a coder despite working with code as part of their work
  3. Struggling to find community and knowledge about how to do their own work

These themes we translated into four core areas of focus which we used to refine ideas and projects:

 

  1. Understand – Reach out and listen to communities to understand their needs, aspirations and incentives and keep iterating to meet these needs at they change. Additionally identify research done on Open Ecosystems and apply it if applicable
  2. Educate – Provide materials, resources and instruction in Open Work including guest lectures, online support and in person assistance
  3. Enable – Make starting to contribute to Open Work effortless and clear. Identify starting points for Open Work, tools needed and help get people started and best practices
  4. Empower – Create raving fans of Open Work that accelerate it’s adoption and  drive the program forward in a sustainable way with their collective vision.

The First Six Months

One of the early opportunities to both understand and educate was through reaching out to Faculty with the goal of guest lecturing in their classes. This provided vital student conversations and also showed professors that we were as willing to contribute as well. This allowed us to talk to students, experiment with learning modules   Referrals to new people grew constantly to the point that we designed a CRM (customer relationship manager) to track all the people we had talked to. After six months we had talked to over 90 people across 29 institutions outside of UVM, and to staff or faculty at 7 of the 9 colleges at UVM. 

 

 

 

Originally we had planned to get a website up right away but that dropped in priority as it became clear that we needed to understand the audience. The final approach here was to create learning tracks that someone could start from on the home page to meet their experience level. 

 

 

Interview

Of 150+ people emails we were able to get 7 interviews (at the time school was not in session).

 

We used  a recorded Zoom session for all interviews with two people from VERSO and a staff or faculty member following a script. Calendly was used for scheduling interviews. Due to difficulties in getting a list of emails we used the public facing report of salaries to get a list of staff and faculty, then reverse engineered email addresses in a google sheet. By using a random number generator we created random groups and send out emails in batches one group (~30 people at a time). This was during the summer so there was a low response rate. We offered a gift card in some situations if it helped provide motivation to join a call, but not across all interviewees.

Supporting Materials

Interview Script
Email Template

Class Lectures

Early on we started reaching out to professors for classes that we found might have potential links to Open Work in order to come in as guest lecturers. This serves both to give direct exposure to the most likely populations to be interested in VERSO, but also builds relationships within the faculty that have already proven invaluable. It also demonstrates to the students that we are here to provide value to them, not just ask for their time and energy.

 

  • ME 265 / BSAD 293 Integrated Product Development (Erik Monsen) – Sept 15 
  • CDAE 060 Design Innovation I (Sarah Williamson) – Oct 3
  • CS 120 QR: Advanced Programming (Lisa Dion) – Oct 25
  • CDAE 395: Ecological Economics Applications (Joshua Farley) – Oct 25

While VERSO was mentioned, the majority of the time was focused on teaching the content for the class by leveraging our industry and academic experience and connecting it throughout with Open Source work and methodology. 

 

These class engagements have also been useful in providing feedback into what concepts and terms students are familiar with, and what they expect out of the student-teacher relationship. We adopted a co-teaching method with both John Meluso and Kendall Fortney teaching the classes together. Modules were designed to be engaging and active learning processes utilizing group problem-solving activities and Think/Pair/Share to surface student insights. The secondary goal is to refine, reuse and specifically share these modules with other OSPO’s to use in their program if they want to. The method of sharing is going to be facilitated by OSPO++ through the workshop series we have been attending. 

The Engineering Capstone (SEED) pushed engagement to the next semester given how far they were into planning the course, but the goal is to do a week-long design sprint with the students when that time comes.

Teaching Modules

External Event

It was clear from the beginning that networking with other OSPOs and Open Source focused organizations was going to be key, but it was not until September that events brought us together with UC Santa Cruz, RIT St. Louis University and Carnegie Mellon. In each of these events we were on panels talking about academic OSPOs:

  • RIT Open Symposium – Sept 7-9 
  • OSPO++ Guide for OSPOs in Academic Institutions Workshop – Sept 13 
  • UC Santa Cruz – UCSC Open Source Symposium – Sept 27
  • Synaptic Supercollider Conference –  Oct 1
  • OSPO++ Skills Workshop – Oct 3
  • Burlington Data Scientist Meetup Open Data – Oct 27

These conferences also served to connect VERSO to other organizations like Red Hat, Amazon, NASA, IEEE and more. There were 30+ new contacts from these events and they spurred the future OSPO Quarterly Planning event by OSPO++ where all the academic OSPOs meet to share what they are working on and build a strategy for sharing work like teaching modules, etc. 

We have attended several OSPO++ workshops focused on defining and supporting the role of academic OSPOs including the workshop in Dublin (attended remotely) to help define some of the needs and duties of the program. 

UC Santa Cruz’s Symposium brought many of the same people from RIT back together. We were on a panel about academic OSPOs and developed interesting thoughts on the role and boundaries of the office.

The Burlington Data Scientist Meetup is a joint presentation by Professor Jean-Gabriel Young and Kendall Fortney to talk about what is open data, where you can find it and how to go about creating it. The audience is the general public of Burlington and UVM with usually 50 or so participants.

We are planning to have a conference next year at UVM to bring together the local open source community into UVM to interact with the students. There is an intent to fund the conference with a grant and potentially follow the Unconference model. The goal is to grow the connection between UVM and the local community to foster new partnerships.

 

Student Organizations

There are over 290 student clubs on campus with a constant collection of events but most are not likely, at this point, to have overlap with the Open Work space. Professors and Lecturers have provided recommendation of clubs to reach out to including:

  • Cat Coders
  • CSCrew
  • SWICS
  • Design Club
  • Entrepreneur Club

There is a plan evolving to bring a joint session of the Design Club and the Entrepreneur Club to talk about open innovation and the development of more support for that kind of work.

Curriculum

Like RIT before us, we identified early on that a formal academic structure including classes, minors, grad certificates, continued learning certificates would both help to provide internal legitimacy, provide a new revenue source, expand the audience and drive the program by growing a group of VERSO alum that have been part of our programing. 

We have started designing an Open Source Course for  Computer Science. We have a framework with a proposal and are working on getting it certified so we can launch it for Fall 2023. It will be an online course with pre-recorded lectures and course content.

This is meant to be the first step in creating an expanded academic catalog that can offer value to students that are interested in Open Work and finding it useful for their resumes to demonstrate their familiarity. It has the potential to expand to include a minor, grad certificate and extension certificate once value has been demonstrated.

Events

It was identified early on that events would be key in building community, driving awareness and providing education of Open Source and that planning and execution of events could not proceed until research was done to understand the needs of the students, staff and faculty and we had people interested in helping. To that end we have been building a set of potential events, identifying partners, getting initial agreements and logistics determined but have yet to schedule an event. Here are a couple of proposals:

  • Humanitarian Open Street Mapping Event – This brings back something that use to be at UVM, in conjunction with Code for BTV and open to the general public
  • Art & AI Symposium – Jann Karson, with Burlington City Arts, potential to include parts about open models and ethics
  • Open UnConference – The Open UnConference (Not Scheduled, likely Fall 2023) – A conference focused on Open themes like Open Source, Open Access, Open Design using unconference structures (maybe) and needs a grant to run
  • PRISM Event – Working with the Prism group to create an event tailored to their needs, they are still defining what that need would be
  • Hackathon – Suggest but details are not fully written out, needs a grant to run
  • Open Source License Workshop  – Focused on Faculty to help them understand what licenses mean and how to attach one to their work
  • VERSO Lecture Series – Bring in external community members to talk about how they use Open Source in their work

Open Access

The importance of open research became quickly apparent and early on we started exploring what the current state was by meeting with Scholarworks, the UVM Pre-print Publishing site to identify the process with which research is currently publicly available. 

All undergraduate work is published on Scholarworks, but faculty adoption is scattered at best. We were able to get some basic metrics for Scholarworks with the goal of measuring our ability to increase adoption. We do not want to create mandates but rather promote best practices and make open access effortless.

There are services like Netfiles and Webfiles exist for internal and external collaboration but there is no UVM data portal for research data. We started to work with the IT team on their planned expansion of VACC (Vermont Advanced Computing Core) to include a public portal which is scheduled to launch in the fall.  In reality there is a potential in the idea of creating open data infrastructure to support faculty and potential drive research value.

Additionally we talked to one of the labs focused on geospatial data and started a conversation to create an ArcGIS hub for all data that can be made public from their research. The next steps are getting permissions and connecting a hub that can centralize what already exists. The size of these datasets are quite large and present technical challenges to provide access. 

The support of open journals has also emerged as a need. The UVM Library talked to several journals about hosting them on the UVM Press platform as long as they are Diamond level open access.

While the details and funding have not been fully fleshed out, we are also exploring the idea of an VERSO Open Access Award to recognize faculty who lead UVM in publishing their articles, code and data in an open and accessible way. The challenge is how to determine what research is published where by whom in a scalable way. 

Research

VERSO has both a Graduate student Jonathan St-Onge and Post-Doctorate John Meluso funded as part of the program.

Relevant academic publications and pre-prints from our team on open source ecosystems:

  1. Trujillo, M. Z., Hébert-Dufresne, L., & Bagrow, J. (2022). The penumbra of open source: projects outside of centralized platforms are longer maintained, more academic, and more collaborative. EPJ Data Science, 11(1), 31. [link]
  2. Warrick, M., Rosenblatt, S. F., Young, J. G., Casari, A., Hébert-Dufresne, L., & Bagrow, J. (2022). The OCEAN mailing list data set: Network analysis spanning mailing lists and code repositories. arXiv preprint arXiv:2204.00603. [link]

Projects

Controlled Digital Lending (here)

The first large project was Controlled Digital Lending (here) which is a partner project with the Boston Library Consortium. The goal of the project is to facilitate the lending of digital copies of materials with copyright constraints across multiple institutions in the Consortium. ReShare is a technical open source collaborator that is supporting the software for requesting a loan. We facilitated a week-long design sprint to help the group clarify the workflow, identify risks and scope the minimal viable product for the first iteration.

Open Source Grant Resource Library

We also identified a need for an Open Source Grant Resource Library through our interviews. UVM Research Development (https://www.uvm.edu/ovpr/research-development) is a small group of people that help faculty understand and apply for grants. They identified the need for templates to explain the context, format and needs for a given grant and started building out a structure to support it, but asked our help in designing the project to be as open as possible (the structure could be shared outside of UVM, the content could not). This could include potentially bringing in faculty to share samples of grant sections as inspiration to help others when they apply.

Allotaxonometer

An internal project is focused on translating the Allotaxonometer (https://gitlab.com/compstorylab/allotaxonometer) from Matlab into Python or Javascript to facilitate better adoption and access. This project is in the initial scope stages so details are likely to change.

UVM Compute project

We are restarting the UVM Compute project (https://github.com/UVMComputes/uvmc-api). It creates super computers from spare compute power from people using websites (much like SETI) and could be quite impactful once deployed after a little more work. Funding and next steps are in progress.