PyCon 2025 Thursday - Tutorial Sessions

When I went to EMNLP last year a coworker who attends conferences frequently gave me some advice: "TAKE NOTES! There will be so much content you will not remember it all". A nice upside of this is I now have notes a can share here.

I attended PyCon 2025 in Pittsburgh this year since my employer was a sponsor. Thursday was a tutorial day and I got to attend two different sessions.

The first session I sat in on was Event Sourcing From The Ground Up. The code for the tutorial is available in a public repository. Overall I really enjoyed the structure: we had alternating periods of lecture with individual hands on time to solve problems. Sebastiaan also provided code checkpoints so that after each section you could build off their example code rather than continuing with your own.

Event sourcing is tricky though. The happy path scenarios are all very encouraging. However when you dig in more you see the complexity unfold. Unfortunately there is only enough time in a tutorial to cover basic pieces so you don't get to see how complex it would be in practice. I should say to that I went into the tutorial with some bias against event sourcing due to having seen it misapplied at my first job with investment banks. The phrase "turn the database inside out" still haunts me.

Regardless I'm still happy I attended, it was fun! Also Sebastiaan recommended the book Architecture Patterns with Python and I'm keen to give it a read. If you are interested in the topic the repository is set up in a way that you could follow without an instructor.

The final tutorial of the day that I attended was Handcrafting Your Own Language Processor: The Art of Writing Recursive Descent Parsers. This session had a repository however it started off empty. Instead Ariel periodically pushed changes to the repository as a reference. Additionally there was a site with instructions that we all followed as we went. I think I preferred the style from the previous session where we had individual time rather than following the instructor as they went. I still got a lot out of it though. If you wanted to try this out the repository would be very useful but the site is very well put together and should be more than enough.

Finally I really wanted to go to Polars: Faster, Simpler, Smarter Python Analytics but it reached capacity. I had a coworker that went and says it was quite good. There isn't a public repository or site for the session but the speak is the author of Effective Polars and has a code example repository.