Last week, I took a lovely train ride to Munich, Germany, to represent GNOME at the Google Summer of Code Mentor Summit 2025. This was my first time attending the event, as previous editions were held in the US, which was always a bit too hard to travel.
This was also my first time at an event with the “unconference” format, which I found to be quite interesting and engaging. I was able to contribute to a few discussions and hear a variety of different perspectives from other contributors. It seems that when done well, this format can lead to much richer conversations than our usual, pre-scheduled “one-to-many” talks.
The event was attended by a variety of free and open-source communities from all over the world. These groups are building open solutions for everything from cloud software and climate applications to programming languages, academia, and of course, AI. This diversity was a great opportunity to learn about the challenges other software communities face and their unique circumstances.
There was a nice discussion with the people behind MusicBrainz. I was happy and surprised to find out that they are the largest database of music metadata in the world, and that pretty much all popular music streaming services, record labels, and similar groups consume their data in some way.
Funding the project is a constant challenge for them, given that they offer a public API that everyone can consume. They’ve managed over the years by making direct contact with these large companies, developing relationships with the decision-makers inside, and even sometimes publicly highlighting how these highly profitable businesses rely on FOSS projects that struggle with funding. Interesting stories. :)
There was a large discussion about “AI slop,” particularly in GSoC applications. This is a struggle we’ve faced in GNOME as well, with a flood of AI-generated proposals. The Google Open Source team was firm that it’s up to each organization to set its own criteria for accepting interns, including rules for contributions. Many communities shared their experiences, and the common solution seems to be reducing the importance of the GSoC proposal document. Instead, organizations are focusing on requiring a history of small, “first-timer” contributions and conducting short video interviews to discuss that work. This gives us more confidence that the applicant truly understands what they are doing.
GSoC is not a “pay-for-feature” initiative, neither for Google nor for GNOME. We see this as an opportunity to empower newcomers to become long-term GNOME contributors. Funding free and open-source work is hard, especially for people in less privileged places of the world, and initiatives like GSoC and Outreachy allow these people to participate and find career opportunities in our spaces. We have a large number of GSoC interns who have become long-term maintainers and contributors to GNOME. Many others have joined different industries, bringing their GNOME expertise and tech with them. It’s been a net-positive experience for Google, GNOME, and the contributors over the past decades.
Our very own Karen Sandler was there and organized a discussion around diversity. This topic is as relevant as ever, especially given recent challenges to these initiatives in the US. We discussed ideas on how to make communities more diverse and better support the existing diverse members of our communities.
It was quite inspiring. Communities from various other projects shared their stories and results, and to me, it just confirmed my perception: while diverse communities are hard to build, they can achieve much more than non-diverse ones in the long run. It is always worth investing in people.
As always, the “hallway track” was incredibly fruitful. I had great chats with Carl Schwan (from KDE) about event organizing (comparing notes on GUADEC, Akademy, and LAS) and cross-community collaboration around technologies like Flathub and Flatpak. I also caught up with Claudio Wunder, who did engagement work for GNOME in the past and has always been a great supporter of our project. His insights into the dynamics of other non-profit foundations sparked some interesting discussions about the challenges we face in our foundation.
I also had a great conversation with Till Kamppeter (from OpenPrinting) about the future of printing in xdg-desktop-portals. We focused on agreeing on a direction for new dialog features, like print preview and other custom app-embedded settings. This was the third time I’ve run into Till at a conference this year! :)
I met plenty of new people and had various off-topic chats as well. The event was only two days long, but thanks to the unconference format, I ended up engaging far more with participants than I usually do in that amount of time.
I also had the chance to give a lightning talk about GNOME’s long history with Google Summer of Code and how the program has helped us build our community. It was a great opportunity to also share our wider goals, like building a desktop for everyone and our focus on being a people-centric community.
Finally, I’d like to thank Google for sponsoring my trip, and for providing the space for us all to talk about our communities and learn from so many others.