2020#61 Readings

Finished that post, now started the next. And having all Fridays off is awesome.

📯 The RDD paper: introducing the Spark general purpose framework

I managed to finish my summary of the foundational RDD paper. I hope you like it.

An introduction to Data Oriented Design with Rust

Micro-optimisation is always the last resort, but sometimes you have to do them. And in this case, knowing this kind of internals is what saves your bacon from frying on the CPU.

Venn Diagrams with D3.js

This has been handy. It’s mathematically fascinatingly complex to calculate.

Starduino – 3D Gaming in 28KB

I have had an Arduboy for a couple of years already (I just wrote a dumb Mandelbrot set renderer, but I can’t find the code or pictures). I ignored it for a long while and then found out the amazing new games available for it. I have played Starduino, and it is impressive, even more after reading the details here.

AI has cracked a key mathematical puzzle for understanding our world

This is a very clickbaity title, for a minor breakthrough in computing states of PDEs using deep learning. I’m not sure what the fuss is: deep learning and NNs in general is high-dimensional function approximation. Solving PDEs is, too. So? Probably the most awesome part is that MC Hammer retweeted the related paper.

The wonderful world of Chinese hi-fi

Interesting! I’m not a particularly high level audiophile, but a cheap set of high quality earbuds? That sounds most excellent.

The Prominent Tech Team Leadership Styles of Middle Earth

This made me smile.

How will the pandemic impact software programming jobs?

Meanwhile, people who know what they are doing and can be relied to work well from home are going to be more in demand than ever.

Raspberry Pi 400 - First Impressions

I have too many unused Pis, but I probably will get this one as well.

DuckDB – The little OLAP database that could. TPC-DS Benchmark Results and First Impressions.

DuckDB is to OLAP what Sqlite is to OLTP. I think it is always important to have small tools that can replace large tools, for use cases when you don’t need the full strength of the tool.

Designing Raspberry Pi 400

More information on the RPi 400, which may be the ideal Christmas gift.

Data Quality at Airbnb. Part 1 — Rebuilding at Scale

I’m stealing borrowing a few ideas from here.

Is this Mahler? This sounds like Mahler

A neat project. I’ve had an inky pHat for 3 years and still done nothing with it 😕.

How to Improve the Spark Exclusion Mechanism in Databricks

Seems like a very reasonable incremental improvement. Marking nodes for decommissioning when failing tasks sounds reasonable.

The Digital Nomads Did Not Prepare for This

A NY times piece that reads a bit like “America’s Worst Nomad Workers”. Chuckled out loud at this one:

[…] went to the immigration bureau and asked for political asylum.

“I said, ‘Trump’s a dictator, my city is burning, and people are dying,’” he said, citing the president, protests against police violence and the virus. “They made a joke that I was the first person since the Vietnam War from America to ask for that.”

🍿 The Art of Code - Dylan Beattie

An entertaining talk by the author of the Rockstar programming language, worth watching only to get to the grand finale. Some minor gripes: the basis for the study of iterated complex polynomials (and functions) was laid by Pierre Fatou (our first cat was named Fatou 😿) and Gaston Julia (born the same day as me), and the first image of the Mandelbrot set was by Robert Brooks and Peter Matelski. I used to work in this field, so these stood out as sore thumbs in an otherwise excellent talk (so for people not in the area, it’s just an excellent talk!).

Buy me a coffee

• Rust
• d3js
• ReadingsOfTheWeek
• Readings
• Apache Spark
• Data