A Community-First PGConfEU in Riga
PostgreSQL Conference Europe 2025 landed in Riga, Latvia, this October, and for returning attendees the shape of the event had shifted noticeably. The traditional pre-conference training day was replaced by a community events day, which by all accounts drew positive feedback. I did not attend it myself, but the word from those who did suggests it is worth factoring into travel plans next time.
Keynote and Session Highlights
The main conference opened on Wednesday, 22 October. After a lively welcome from Magnus, Karen Sandler delivered the keynote, tying her personal history with a heart condition to a broader argument about why free and open source software matters today. For many in the room, the discussion of copyleft and software freedom offered a useful reminder of the legal and ethical foundations under the code we work with daily.
Sessions ran across four parallel tracks, organized by theme: app developer, PostgreSQL internals, DBA, community, and sponsors. That division made the perennial choice between simultaneous talks a bit easier to navigate by interest area.
A few talks stood out across the range of topics covered. Laurenz Albe's session on common PostgreSQL anti-patterns and how to avoid them was both well-prepared and entertaining. Melanie's talk traced the journey of a patch to improve the freeze feature in VACUUM, from initial idea to commit, offering an honest look at the development process, including the need to redefine problems mid-way and accept failed attempts.
Two sessions covered pg_tde from different angles. One explored the motivations behind the extension and how past community work helped shape the final solution; the other dove into the technical workings of pg_tDE and what is planned next. Polina's talk on the hazards of logical decoding shared practical lessons on using replication correctly, while a DBtune session addressed AI-driven performance tuning for PostgreSQL flavors.
The overall selection felt strong this year. Topics ranged from finding your entry point into PostgreSQL contribution, including for non-technical people, to explaining PostgreSQL to kids in an accessible way. A talk on fast-path locking in PG18 was particularly notable, borrowing the concept of associative sets from CPU architecture to speed up locking. Another presentation, from Antonin and Alvaro, showed how the table repacking logic from pg_squeeze is making its way into core, a demonstration that solving a problem properly can lead to broad adoption.
Panel on Neurodiversity
A panel discussion addressed how PostgreSQL community members can be more aware and accommodating of neurodivergent colleagues and contributors. The conversation was informative and a genuine eye-opener, illustrating how small choices in communication and environment can either meaningfully help or actively hinder someone else's participation.
Looking Back and Ahead
Returning from another PostgreSQL conference, the lasting impression is the breadth of work happening both with and inside PostgreSQL. Between the technical depth of the sessions and the concentrated chance to trade ideas with so many engaged people, the event continues to be a reliable source of new perspectives and renewed energy for the work ahead.
