HomeNews and blogs hub

Geo+Code UK - Lessons from organizing sprints for a mixed audience

Bookmark this page Bookmarked

Geo+Code UK - Lessons from organizing sprints for a mixed audience

Author(s)
Leonardo Uleda

Leonardo Uieda

SSI Fellow

Posted on 29 May 2023

Estimated read time: 7 min
Sections in this article
Share on blog/article:
Twitter LinkedIn

Geo+Code UK - Lessons from organizing sprints for a mixed audience

Posted by d.barclay on 29 May 2023 - 11:30am Geo+Code participants around a large screen showing a demo of open educational resources using JupyterBookPhoto of Geo+Code participants around a large screen
showing a demo of open educational resources
using JupyterBook.

By SSI fellow Leonardo Uieda.

After moving to the UK in 2019, I found myself missing some of the connections I had made in my previous job with folks that live on the border between geoscience and open research software. So I based my SSI Fellowship around establishing new connections within the UK. Or at least that was the plan. Reality ended up being 2 whole years away from in-person events because of the COVID19 pandemic and very little time and energy to devote to socializing.

When 2022 rolled in with relaxations in social restrictions, I was thrilled to finally be able to organize an event for myself and my fellow geo-coders. That’s how the “Geo+Code” event came to be. My main idea was to gather UK geoscientists who code for a 1-day event in Liverpool. The format was inspired by the “sprints” that happen after events like the Scipy Conference, where everyone gathers to work on coding projects for 1-2 days. The format is great for building relationships and producing something concrete as an output.

This blog post reports on how the event went, some of the challenges we faced, and some of the things that worked well.

The event

After brainstorming some ideas, I came up with the following goals for the event:

  • This would be a gathering of people in the intersection of geoscience, coding, education, and open source/data/science.
  • The focus is on doing something together that can be the start of great collaborations going forward.
  • We’ll gather for 1 day, in-person, at a nice venue.
  • Participants will have the chance to introduce themselves and pitch a project idea.
  • All communication will be through the Software Underground Slack, an awesome open community. Having everyone there ensures that conversations can easily continue after the event.
  • During the event, we’ll pick one or more projects to work on together.
  • The projects could be coding something, producing a video, joining teaching material, or anything generating open resources.

I put out a call for participants in the Software Underground and got several responses. One of the first was from 2023 SSI Fellow Michael Tso, who was instrumental in getting the event organized and the venue sorted (thanks, Michael!). On 28 November 2022, 14 participants from 10 different institutions showed up at Blackburne House, Liverpool, for a day of chatting, coding, and content creation.

The event was a success, with good feedback from participants and a lot of happy faces. Most participants didn’t know each other previously or had only met online. During the event, we settled on two projects:

  1. Reviewing the tutorials and examples in the documentation of SimPEG, an open-source geophysics software. This group had the objective to comb through the examples and identify any broken code or sections that required more detailed instructions in the software. A thing to note is that none of the participants is a core developer on SimPEG.
  2. Developing an open textbook for applied geophysics that can be used to teach undergraduate courses. The book would be based on JupyterBook and would include computational lessons and open data. The book would be hosted under the Geophysics Library GitHub organization.

Challenge: Organizing an inclusive event

From the start, I wanted to make sure that this event would be as inclusive as possible and included participants from diverse backgrounds. Unsurprisingly, putting out a call for volunteers did not produce a very diverse set of attendants. So I explicitly invited people who would be a great fit for the event but wouldn’t necessarily put themselves forward as participants, which was much more effective at attracting participants than simply putting out an open call.

Since other participants and I have young children, I tried to schedule the event for a weekday and provide childcare support for those who would require it. This proved to be a major difficulty. Local nurseries wouldn’t accept children for a one-time event and it was difficult to find a provider who would be able to care for 1-5 children only for a single day. In the end, none of the participants ended up needing the childcare and I have yet to find a good solution for this.

Challenge: Variable levels of technical skill

Our participants came with a range of technical backgrounds, some were very experienced in software engineering and others only familiar with the content of Software Carpentry lessons. This meant that a single day of working together on a shared GitHub repository was not enough. We spent most of the day getting everyone up to speed on:

  • Setting up a working development environment with conda and git.
  • Forking and cloning a GitHub repository.
  • Creating branches and pushing them to GitHub.
  • Opening and merging pull requests.
  • Building HTML from Jupyter Notebooks using JupyterBook.

Between this and brainstorming what we wanted to include in our book, we were left with very little time for actually creating content for the book. By the time we wrapped up, our book was little more than a test document with people’s names and a skeleton for the chapters.

Ideally, this should have been organized as a 2-day event, with training on day 1 and sprinting on day 2. However, making the event longer conflicts with its inclusivity of it because it requires a higher commitment than some may be able to make. In the future, I hope to experiment with a hybrid model where the training is done online and async ahead of the in-person event.

Outcomes

The outcomes of the day were:

  1. Participants identified that the main source of frustration with the SimPEG examples was obtaining a working environment with the software installed. A considerable amount of time was spent trying to get a working version of the software environment on everyone’s computers. Afterwards, the examples themselves were executed without any issues. This would be fed back to the SimPEG developers as something to focus on in the future.
  2. The open applied geophysics book was started at https://github.com/GeophysicsLibrary/applied-geophysics. The scope and main direction for the book have been laid out already and hopefully, a future event can be used to kickstart the content.

It was also great to meet so many new people, most of which I only knew virtually (or not at all). Everyone seemed to come out of it excited and some were even considering hosting a follow-up event at their institutions in the future.

I hope this post can be helpful for those who want to organize similar events. Of course, I’d be more than happy to get suggestions for how I could do things better in the future. Drop me a message!

Share on blog/article:
Twitter LinkedIn