Before the conference

My partner and I arrived in Athens a few days before the conference. It was a convenient excuse to visit a European capital we hadn’t been to yet, and of course to eat as much delicious food as possible.

One day before the conference I went to the django.social event organized by Jon Gould and Andrew Miller.

The bar they initially picked got too crowded, so the group moved to the backup place. I came late to the first bar, found no one there, and was a bit lost until I made it to the right spot. Once I did, I felt immediately at home. I met old friends, made some new ones, and had a very fun evening. So fun, in fact, that I forgot to take any photos, so I have no social proof that I was there 😅

Talks

The next day the conference started. Time to listen to some great talks and mingle with everyone during the breaks. I enjoyed pretty much all the talks and have pages of handwritten notes. A few of the most memorable ones:

• Static Islands, Dynamic Sea by Carlton
• AI-Assisted Contributions and Maintainer Load by Paolo
• Digitising Historical Caving Data with Python and Django by Andrew
• Body of knowledge by Daniele
• Auto-prefetching with model field fetch modes in Django 6.1 by Jacob
• Where did it all BEGIN;? by Charlie and Sam

Lightning talks

Lightning talks are my favorite part of any conference and they didn’t disappoint at DjangoCon.

I also gave a lightning talk of my own on the last day. It was a shorter version of my talk from the Python Lisbon Meetup earlier this month on how we sped up Django startup times with lazy imports.

Sprints

For the sprints I decided to see if our work project (Fencer) works on the main branch of Django. There are many features I’m looking forward to, and I wanted to make sure we’ll be in good shape when the final release lands.

Regressions

Our test suite unearthed two regressions:

1. #37047 Crash in Query.orderby_issubset_groupby for descending and random order_by strings
2. #37048 Backwards incompatible change to InclusionAdminNode

The first one was fixed by changing the code to handle the edge cases of descending and random sorting. My PR with the fix was merged during the sprints thanks to Jacob!

The second ticket was initially resolved as wontfix since the regression was in an undocumented API. We did end up merging a note about it to the changelog. I also got my fix merged into django-unfold, which was what led me to the issue in the first place.

Fetch modes

With the above issues fixed, our codebase ran smoothly on Django 6.1 and I was able to play around with the new model field fetch modes. From what I can tell, they work perfectly. I’ll write a longer blog post about this closer to the release date, but it’s great to see the number one reason for N+1 queries going away soon!

Fin

As I am writing this I am feeling exhausted. As an introvert, all the conferencing drained my batteries. Time to retreat home and recharge until the next one (PyCon Portugal in September). Which reminds me: I need to submit my talk proposals. The deadline is April 30th AoE! Here’s the CFP.

I’m terrible at taking selfies so I stole this one from Paolo, I hope he doesn’t mind! PS: he also wrote his own writeup on DjangoCon 2026. Be sure to check it out!