Our inaugural fall festival was a great success!
Three years ago, I envisioned an online event where our community could come together to celebrate open source and open science, share knowledge, and learn new skills. Last month, that vision became reality with pyOpenSci’s first-ever Fall Festival, held from October 28 to November 1. The event brought together 64 participants from over 15 countries—a global mix of researchers, developers, educators, and Python enthusiasts.
The week was packed with inspiring keynotes, hands-on workshops, and informal office hours, where participants connected and reflected on their learning.
What did you enjoy most about the fall festival:
Getting to know about pyOpenSci as an organization and resource. Also, [being provided with] practical, professional tools that I can use right away.
Acknowledging our amazing pyOpenSci team
Events like the Fall Festival don’t happen without an incredible team working behind the scenes to support every step of the process. I want to extend our heartfelt gratitude to all the instructors, helpers, and keynote speakers who contributed their time and expertise to make this event truly special. Your dedication, energy, and enthusiasm ensured an engaging and impactful experience for everyone involved!
A special shout-out to Carol Willing, Jeremiah Paige, and Jonny Saunders, who supported multiple workshops and co-developed and reviewed many of the lessons now published online. We’re so lucky to have such an incredible community. 🫶
What made this event special
The vibe made this event special. The energy of participants who wanted to learn together was supported by the vibrant and knowledgeable pyOpenSci community. While many mentioned me directly, the vibe was more than just me. Our community came together to help scientists learn hard technical skills–together and without judgment.
I love the engagement…how people were attended to individually despite being in a group setting. I enjoyed that learning could happen in a personal and group setting.
We met learners where they were at!
The enthusiasm of Leah and her friends, Spatial chat and all new Python things I didn’t used before like Packaging, Great Tables, Quarto, Clean Code and Programming in [GitHub] CodeSpaces
Using the interactive platform SpatialChat rather than a traditional online platform like Zoom helped to create that vibe. More on that below.
Keynote talks
We kicked the event off on Monday, October 28, with a morning of KeyNote talks headlined by Eric Ma, Melissa Mendonça, and Rowan Cockett. The Monday talks aligned perfectly with the training events held Tuesday through Friday. The talks set the stage for a truly engaging week of learning together.
Eric Ma: The human side of clean code
Eric’s talk highlighted how simple practices like clear documentation, readable code, and user-friendly installation can amplify the impact of data science projects.
A standout moment? The “Roast Your Repo” exercise! Eric invited attendees to critique a repository from his thesis, showcasing the power of small improvements—like adding a fleshed-out README or modularizing code—to make research reusable and collaborative. It was a fun, hands-on way to explore how the human touch transforms code quality. And let’s be honest, we all likely have one of those code bases or repos from our early degrees! I sure do (and it’s not even on GitHub!).
Melissa Mendonça: From academia to open source
Melissa shared her journey from academia to open source software development, reflecting on the courage it takes to step into the unknown. She celebrated the scientific Python ecosystem, emphasizing how libraries like NumPy and SciPy enable countless domain-specific projects.
Melissa also highlighted the challenges of volunteer-driven communities, stressing the need for clear governance and transparent user engagement. Her focus on open science principles—transparency, reproducibility, and accessibility—was inspiring and reinforced the importance of collaborative, inclusive practices.
Rowan Cockett: Rethinking scientific publishing
Rowan invited us to imagine a future where scientific publications are as dynamic as the research they describe. He introduced Myst Markdown tools that blend code, data, and narrative and discussed his company, CurveNote, which aims to revolutionize scientific publishing.
His vision of collaborative, interactive, and automated publishing draws from open source principles and challenges the static nature of traditional papers. By rethinking how we share research, Rowan reminded us that we already have the tools to make science more immediately accessible and impactful.
Workshops and Quarto Live
We also invited George Stagg (Posit), developer of Quarto Live and James Balamuta to talk about how Quarto Live is empowering education. Quarto Live makes it easy for you to create interactive data science environments for learning in the browser.
- Yes, it runs on GitHub pages - no servers needed.
- Yes, students can write Python on their phones.
Both talks are below.
George Stagg
James Balamuta
Let the (open science & open source) learning begin!
The main “course” of the event was 4 days of active learning. Here, participants engaged in interactive workshops designed to build skills and confidence in writing, sharing, and publishing scientific code. Using SpatialChat, we created a dynamic and collaborative environment that fostered real-time learning and group discussions—something you just can’t replicate in traditional video call platforms like Zoom.
Our curriculum followed a clear narrative:
Write better code → Package it → Share it → Tell an interactive data story.
[I enjoyed….] Leah’s teaching style, the platform, and the tutorials. I felt very ease learning together with the pyOpenSci community.
Each day introduced new tools and practices to help participants transform their workflows and make their science more accessible, reusable, and impactful. Here’s how we approached it:
Day 1: Write Better, Cleaner Scientific Code
We started by diving into writing better, cleaner code. Participants explored how to make their code:
- DRY (Don’t Repeat Yourself) and modular,
- Expressive and easy to understand,
- Consistent with Python style guidelines like PEP 8.
We introduced strategies to improve robustness, such as creating functions and using tests and checks to validate outputs.
Day 2: Create Your First Python Package
On Day 2, participants learned how to turn their code into reusable, installable Python packages. I’ve taught Python packaging with Hatch several times over the past year, and everyone consistently loves having an all-in-one tool! They also pick it up quickly, making the tool accessible and usable–a win-win! This was one person’s response to their favorite day:
…I thought Hatch was really useful and I liked the fact that it can do a couple of things as one tool.
The packaging day covered:
- An overview of Python packaging for beginners,
- Making code installable in Python environments with tools like Hatch,
- Adding package metadata using a pyproject.toml file,
- Including a license to define how others can use and share your Python package.
Day 4: share your Python code (with everyone!)
On Day 4, we empowered participants to share their work more broadly. Many learners enjoyed this day which was cool for me because it was entirely new content I had never taught before but thought was so important to any scientist building software and writing code. Below, we asked them what day their favorite was. The theme of this response was common in the feedback:
[I enjoyed] making package installable via PyPI by using hatch (one often writes code, but never gets to this stage)
Key share your code topics included:
- Why sharing code matters,
- Adding a DOI to code using Zenodo,
- Publishing through JOSS or the pyOpenSci peer review process,
- Publishing code on PyPI using Hatch.
- NOTE: we also have a publish to conda-forge lesson.
Day 5: Interactive data storytelling with Quarto & GreatTables
We wrapped up with an introduction to Quarto, a powerful tool for creating dynamic, interactive scientific narratives. Participants explored how to integrate code, data, and findings into a cohesive story—transforming static publications into living, engaging documents.
We also showcased Quarto Live, which lets users dynamically interact with code in the browser. For educators, this opens exciting opportunities to create lessons where students can learn directly in a live coding environment. How cool is that?
Fun fact: GreatTables has been accepted by pyOpenSci with the plan to be fast-tracked through the Journal of Open Source Software (JOSS) after the pyOpenSci review through our JOSS partnership.
Reflections on the 2024 pyOpenSci Fall Festival
The 2024 Fall Festival was an incredible learning experience for pyOpenSci! One of the standout successes was our last-minute switch to using Spatial Chat as our virtual platform. Participants loved how easy and intuitive it was to use, and it quickly became a seamless way to foster interaction and collaboration.
I liked the use of software like spatialchat, where you could break out into groups to work on things. This makes everything a lot more casual and fosters networking.
We were also thrilled to offer 16 scholarships for this event, making the festival accessible to diverse attendees. We’re excited to continue improving our events to maximize engagement and accessibility for all community members. We look forward to building on this success for future festivals!
What’s next for pyOpenSci
Most of the resources used to teach are published on our pyOpenSci lessons website. We’re actively working on updates to incorporate additional content still housed in Google Docs, ensuring it’s accessible to all.
Looking ahead, we’re excited to run more events like this to support pyOpenSci’s broader mission of supporting the open source software scientists need to make their work open. We are planning to:
- Develop a new set of lessons that support using GitHub collaboratively and empower contributors to make their first open source contributions.
- Run online beginner-friendly sprints that further engage the community in making contributors to open source, no matter how big or small, we want to ease pain points!
- Attend and run events at PyCon 2025 and SciPy 2025, holding events and connecting with you!
Check our blog and events page for upcoming events and opportunities to engage with the vibrant pyOpenSci community.
Support open science: Get involved with pyOpenSci
- Volunteer to be an editor in our peer review process
- Submit a scientific Python package to pyOpenSci for peer review
- Donate to pyOpenSci to support scholarships for future training events and the development of new learning content.
- check out our volunteer page for other ways to get involved.
You can also:
- Keep an eye on our events page for upcoming training events.
Follow us on social platforms:
If you are on LinkedIn, check out and subscribe to our newsletter, too.
Leave a comment