2021#22 Readings
4 minutes read | 674 words by Ruben BerenguelA 2-week’s worth of readings means a longer than usual list, as usual.
Last week I saw my parents for the first time in a year and a half.
A Bayesian Analysis of Lego Prices in Python with PyMC3
This is a very long post, and it put my rusty data science abilities to test. It’s an interesting analysis, maybe biased because I like Lego and Python and PyMC3.
Understanding LATERAL joins in PostgreSQL
A feature I have never used. And probably never will, but it is worth knowing.
My (career) investment thesis on dbt Labs
I’m still not convinced DBT is the solution but it certainly is miles better than “just write SQL and pray”.
We Are All Scutoids: A Brand-New Shape, Explained
Designing Beautiful Shadows in CSS
Very informative, and the results look awesome.
Adoption of engineering standards at Typeform
I can confirm that standards can be proposed, discussed and that adoption is smooth!
đ Learning How to Learn
I stumbled upon this book when looking up material about concept maps, in turn this is a tool I found in G. Klein’s (et al) Working Minds: A Practitioner’s Guide to Cognitive Task Analysis to share expertise/knowledge on a system. I found concept maps interesting enough that I diverted into checking what information about them is available. I’ll write some of my findings (with examples) in a future post.
Open-water races and our professional development
If you don’t swim, the analogy is similar with running and cycling. The theme about energy management is spot on.
Writing a book: is it worth it?
From what I know from other tech authors, “breaking even” is almost impossible.
Treating data as a product at Adevinta
Probably the separation in domains is the most widely usable part. Worth reading.
DIY RGB Icosahedron build
Wow
Structural pattern matching in Python 3.10
This is an in-depth analysis of what pattern matching in Python 3.10 looks like, with examples of use applied to real code. He’s a bit on the negative side, particularly given that this feature appears after Python becomes 30. From my point of view, it’s a great feature to have, but that’s seen from Scala. Dataclasses, structural pattern matching and Mypy make Python 3.10+ code look like Scala 3 when using significant indentation in Scala.
đŚ Thread by @shahidkamal
A personal story about the beginnings of a software career, at the dawn of British computer gaming history.
Journal Intime 1 Vera Molnar
Vera MĂłlnar is one of the pioneers in generative art, and here you can find his first “creation notebook” scanned. Here’s the second one. Truly inspiring.
Implementing a Generic Filter Function in Go
I have a personal toy data project I may write in generic Go (this brings memories of having  “Java” and “Generic Java”⌠and actually Phil Wadler was involved in Generic Java and in Go genericsâŚ).
Real-Time Exactly-Once Ad Event Processing with Apache Flink and Kafka
It also uses Apache Pinot, in one of the first blog posts I see where it’s heavily used. Probably a project worth revisiting.
A Perfectly Cromulent Word
Embiggen your vocabulary.
Reflecting on Four Years at Databricks
It’s not the first time I share Haoyi’s programming blog. It’s interesting to see how Databricks has grown and evolved for 4 years.
