2021#27 Readings
6 minutes read | 1074 words by Ruben BerenguelChristmas edition!
Among other things, I got the wonderful Lego bonsai tree (affiliate link) you can find below, and maybe even better: a keyboard hand warmer. It’s like an electric blanket you put under your keyboard and mousepad. I have it on now (at 38 degrees C) and it’s feeling good.
Hard to kill, except for a cat
Native Apple Silicon JDKs
So far I had been using the same version of AdoptOpenJDK we had in production containers (even if it was running under Rosettta…) but I may change to a native one, it’s a significant enough change (close to 2x faster) in speed to be worth it.
Scan of the Month - AirPods Evolution
It’s December’s scan, and the scan has no permalink that I can find, so… Go quickly. It’s a gorgeous site, I like the use they make of scrolling to advance the evolution of the design.
Introducing stack graphs (Github blog)
This is a new code navigation for Python code in Github, based on a way to find call sites and definitions without having to build or fully analyse the code. The approach looks extremely powerful. It got me to treesitter, I was unaware it is a full language parser. I now have a project for it 😏.
Towards Operational Excellence: Mechanisms
A story about Bezos in early Amazon, and the usual story about the stop cord in Toyota. The gist is that having processes that are followed and religiously adhered to wins over good intentions or expecting things to be fixed by magic. Trust the process.
Engineering Fundamentals Checklist
By Microsoft. It’s a very high standard, probably impossible to summarize. Covers way more things that you may consider thinking about. Worth bookmarking and revisiting regularly.
Stacked changes: how Facebook and Google engineers stay unblocked and ship faster
No relationship with the stacks mentioned in the article above.
These large companies use internal code review tools let review PRs stacked on top of (feature) branches. Github notoriously doesn’t, and this post is the launch of a service that adds this feature on top of Github. Interesting, but using it is probably a no-go in most companies 🔐.
Death by coconut - Wikipedia
🤷♂️
3 Common Fallacies About Creativity
Resistance to premature closure — a key aspect of creativity — is our ability to keep an open mind when we already have a potential solution. Some of the best solutions don’t come in the initial meeting or two, but after a longer incubation period.
📚 The Programmer’s Brain: What Every Programmer Needs to Know About Cognition
This is a 6 star book. A bit boring at times, yes, but the information in it should be mandatory knowledge for any developer. Just to give you an idea, I collect notes from many books I read. This one is the one with most notes, and actually has more than the next two with most together.
A comparison of data processing frameworks
This post offers a 2x2 matrix as a way of comparing data processing frameworks, splitting according to being batch, stream+batch, data driven or task driven. This is a good way of seeing these tools. Additionally evaluates the (subjective) maturity of each.
Plato’s Dashboards
How do we know the metrics we use measure anything? An example here breaks metrics into: surrogate variable we measure, how we do it, possible artifacts, what value we are really interested in, and certainty of its measured range. This “framework” would be an excellent way to document metrics.
Management Time: Who’s Got the Monkey?
This is an extremely old article from HBR, but fully valid. It’s about being aware of who has the next task, the manager or the subordinate, using the “monkey in the back” metaphor. From then, it gives a set of rules to take care or feed (or shoot) the monkey.
Manage Your Energy, Not Your Time
A small study on improving the energy of knowledge workers by a set of renewal exercises. They cover the quadrants of body, mind, emotions and spirit. Although these last three may sound the same, they are markedly different. All are known, but having them together makes this a good read.
Introducing our three roles in Data & Insight at Oda
The roles are the expected engineer, analyst and scientist. Specifying the different abilities and skills in the area of data, each role is assigned a “virtual scorecard” (with each skill having a “score”) which defines them.
Why all new managers want to quit
Seems to be a common enough situation that the author wrote a list of reasons why along with an explanation about what each implies in managerial work. I have skimmed some other content from the author and it looks good.
Hallucinating new numbers in a GLSL shader
It combines several new developments together, including a way of converting a trained set of network weights into GLSL code that can run on your browser to create a network that runs in a shader and creates “fake numbers” after being trained on the (I guess) MNIST dataset.
Achieving fast upserts for Apache Druid
This is a terrific post. It explains the Apache druid data hierarchy and query model in enough detail to state the business problem they had, and then how they solved it. Also an excellent example of analysis before choosing a tool/solution.
Density of goals: The secret to high mobile game retention
This is an outlier, because it was a promoted ad in Twitter to teach you how to increase retention in free-to-play games, the horror. But as a player of “premium” games it was interesting to read about goal design, and how RPG mechanics and “open world” games improve player experience.
