London Scala User Group

🎉 Come along to the London Scala Talks! 🎉
In this event you'll hear from Hamish Starling, Paul Stanford and Andi Miller.

Agenda
6:00pm - 🥤 Doors open. Come along and grab a drink!
6:40pm - 🗣️ Hamish Starling: Specialization? In Scala 3?
7:20pm - 🍕 Intermission: Join us for some free food and drinks! Vegan and vegetarian options are provided. Let us know if you'd like something special - we'd be happy to accommodate.
7:50pm - 🗣️ Paul Stanford: Brute Force, Binomials, and Balatro
8:10pm - 🗣️ Andi Miller: Teaching Hedgehogs to Hibernate: Suspendable Query DAGs at Permutive
8:30pm - 🥤 Socialising: Grab a drink and let's discuss the talks.
9:00pm - 🍻 Join us in a pub to discuss the talks!

🌐 This event DOES NOT have a live stream
We hope to see you in person.

🗣️ Hamish Starling: Specialization? In Scala 3?
Generic methods and classes used with primitive types suffer from an unfortunate performance loss in Scala compared to languages like Rust or C++, due to erasure. Scala 2 provided a solution through the @specialized annotation, but until now Scala 3 has been without a suitable replacement. This talk explores the specialization problem and introduces inline traits and specialized traits, two brand new experimental features of the Scala language that aim to bring specialization to Scala 3! We'll discuss what they are and how you can use them to make your code faster!

⭐ Hamish Starling (he/him) ⭐
Hamish is a recent graduate from Imperial, who caught the Scala bug while working on the compiler during his study exchange at EPFL. When he's not thinking about FP you can find him reading about his latest obsession, which could be anything from salad gardens to transit systems...

🗣️ Paul Stanford: Brute Force, Binomials, and Balatro
Balatro's Shortcut Joker lets you make a poker straight where gaps of one are allowed (e.g. A,2,4,5,7 would now count as a straight.) Playing with it the other day I felt like I was getting straights all the time, and I became curious as to how likely that actually was. Perhaps I was a victim of confirmation bias? So, like any counting enthusiast, I decided to work out the exact probability, thinking it would be a fairly straightforward task (pun intended.) I could not have been more mistaken.
I really enjoyed the journey, and I think going through my simple brute force solution and my slightly more nuanced, more mathematical solution might make for a fun little talk! I will assume no knowledge beyond high school maths, and the flavour of Scala that I show will very much be the functional kind, but with a lowercase f.

⭐ Paul Stanford (he/him) ⭐
Paul is a Senior Software Engineer at Permutive and has been a Scala enjoyer since 2019. He has spent his time in the Scala world at assorted startups, from writing graph traversal algorithms that detect financial crime to using data sketches that count huge audiences. In his spare time he enjoys gaming and has unlocked every joker in Balatro.

🗣️ Andi Miller: Teaching Hedgehogs to Hibernate: Suspendable Query DAGs at Permutive
At Permutive, a user's insights query unfolds into a graph of subqueries: realtime ones that return in milliseconds, and warehouse queries that can take minutes. Executing that graph shouldn't mean holding a process, a fibre, and a database connection open until the slowest node finishes. hedgehogs is a small open-source graph library for Scala whose functional visitor concurrently dispatches work as soon as it's ready, sitting midway between an imperative visitor (fast, but mutable) and a catamorphism (principled, but structure-bound), neither of which can hibernate. hedgehogs can: a run's entire state lives in the graph itself, so when only slow external jobs remain it suspends to a database, rehydrating when a callback fires.
Durable execution built from cats-effect primitives, no workflow platform required.