Progress

Well, work goes on.

I've now programmed a small script that does sidenotes automatically for meAnd I'm so happy! And I'll use it extensively, you'll see!. Who knew that parsing parentheses is such an unsolved problem? Well, my script is one big hack upon another, but I don't care.

I needed a CV for a prospective customer. So I did what any self-respecting software developer would do: I started a project for creating CVs.

PlantEd

I'm now proficient enough with Micropython and ESP32s that I can start up things quite quickly. So that's what I did, started some coding for light control and (maybe) moisture sensors. Not very tangible, yet, and no hardware so far.

I also noticed that if I write this notice beforehand, it magically becomes true!

NTS

No progress on NTS this week.

Articles

• The Kingsway Basic License - /dev/lawyer: Interesting comparison between one example of music licensing with software licenses. Also, lots of great license pointers for semi-open licenses.
• All Heroes, No Casualties - /dev/lawyer: A warning, that programming must not only have technology, but also culture.
• Scientific Visualization: Python + Matplotlib, free ebook
• Why you can't solve knowledge problems with information tools alone: Perfect for the recent customer I worked with. It's because knowledge problems cannot be solved with information tools alone.
• Information Is Beautiful: Wonderful visualizations
• What can a Tardigrade “Water Bear” Survive?, from the Journal of Astrological Big Data Ecology!?! (That's satire, by the way.)
• How I helped build a profitable MVP over a weekend
• Goodbye, Clean Code: while the example mentioned is probably not the best example, this is something I've also felt at times: it's good to have a Code for cleanliness, but strict adherence to the exclusion of everything else can lead to worse outcomes. I guess that's true for everything. For programming, I'll stick to "make it work, make it fast, make it nice".
• Ivermectin: Much More Than You Wanted To Know: Lots of analysis on studies of Ivermectin, which does not help. But there is a lot of interesting meta-analysis of the scientific process and on how thinking works. I'm really hung up on this: If you don't know how technology works, but you see technological progress, minaturization, all the marketing promises on AI and "smart" tech -- would you believe it possible to put microchips in vaccines to control/influence people? Is there a qualitative difference between "cars drive themselves", "smart watches that know everything about me cost 20€", "the metaverse", and "computers can be small enough to fit into a medication, yet powerful enough to influence someone"? Unfortunately, I don't think there is, and that's deeply unsettling.
• Evidence of Fraud in an Influential Field Experiment About Dishonesty: Deeply, deeply unsettling.
• Neutron Image Gallery: A Visual Guide to Neutron Imaging
• Mediaeval Mythbusting Blog #3: The Man Who Invented the Spiral Staircase Myth: I love these stories that go like "here's a commonly known fact, but it's wrong". This one is about the chirality of staircases in medieval castles and how it has nothing to do with swordfighting.
• Cracking the Adventure Time Cipher | Aaron Randall
• How to Grow Sodium Chloride Crystals at Home: Growing salt crystals sounds like the easiest thing in the world: make some brine, wait, done. But it's actually quite complicated to make clean salt crystals, such that they are (almost) fully transparent. I'll have to try this soon.
• Toy Project: Recreating the Fortune Telling Minigame from Super Mario Bros. DX
• Tools I use: PyCharm / Idea plugins: I have to try out CodeGlance2, Key Promoter X, Tab Shifter, and definitely need to learn more multi-cursor magic.
• Upgrading the soldered RAM in a Dell XPS13 7390: I am not going to try this.
• Glybera: The Million-dollar drug: Drug is constructed that fixes a genetic defect permanently, but due to market pressure it's not available. However (not in the text): "It doesn't make for a very compelling news story, but the real reason why Glybera failed is because it's not that great of a drug and it loses effect over time and you can't redose with the same AAV vector. They also relied a surrogate endpoint versus actual clinical measures (as those were to confounded by the small sample size and patient variability). It just barely got approved by the EMA and based on feedback the company didn't bother to pursue FDA approval." (via HN)

Libraries, programming, etc

• Browserflow: Browser automation made very simple. Looks nice and useable, but it's tied to a subscription business model (again).
• Duff’s device in 2021: "In conclusion, Duff’s device if still worth considering3, but you’ll have to do your own benchmarks to be sure where it is indeed useful. However, Duff’s device does add more verbosity to the code. Even if it’s easy to document (as stated before), you may not have the time (or not want to spend the time) to do benchmarks and consider Duff’s device. It’s up to you."
• ZipFly: Using this library will save you from having to write the Zip to disk. Some data will be buffered by the zipfile deflater, but memory inflation is going to be very constrained. Data will be written to destination by default at regular 32KB intervals.

Games

• John Wick: Hex: A strategic action game, as strange as that may sound. You plan your actions, and they all use up your time, as does your enemies actions. Would be very nice with a "Katana Zero"-type action replay, just to make one feel super-badass. 7/10
• Eastshade: Very relaxing. Not as beautiful as I had hoped, has more of an 2005-level aesthetic. 8/10
• Craftopia with Anton and bison. Still great.
• Prey, still great there, too.

Recipes

• Ragusea turkey, only slightly adapted for goose. Turned out not super-great, but ok enough. De-boning the goose took about 45 minutes and was not very complicated.

Other media

• Steiner's Porism [SoME1]: Great!
• Adam Savage's First Gundam Build—RX-78-2 Perfect Grade Unleashed!: I like everything about this: I like Adam Savage, I like the builds he makes, and I like this amazing plastic-molded model figure. Well, I'm not into Gundam, but I can appreciate the amazing quality of the model. They run between 300-400 €, by the way.
• Plastic Injection Molding - The Engineerguy
• Marvels of Injection Molding - This Old Tony