Pi Day is one of the most important days for bears on this island, only second to Christmas. The day to celebrate depends on whether you pick the Japanese date (14 March, "３月１４日") or the British date (22 July).
I started making things for Pi Day in 2020, and have been doing so every Pi Day (Japanese date) after. Unfortunately, I never took the time to tidy up my projects. As a result, some of them end up on GitHub, some on BitBucket, some on Cloud only, and some only in videos, source code completely lost. At the time, I also took the opportunity to try "new technologies" to me. That was rewarding, but it also meant it was almost impossible to put them together at one place to show cubs.
But these days, it seems WebAssembly is quite mature now, so I should be able to put things in various languages on a web page. To be clear, I attempted Python on WebAssembly last year, but because of the quirks in Svelte (what I used to build my previous blog), I could not use PyScript and had to use Pyodide directly, plus styling a notebook was a lot more work than what I expected.
Both problems seem to have been solved by the Jupyter Lite project. I have yet tried it, but at least there's a possibility. So I dumped everything I did in previous years here and will probably remake some of them this year.
| Year | Theme | Content | Technologies |
|---|---|---|---|
| 2020 | Wave Slides | Sine and cosine, AM and FM, Fourier Series, Lissajous | WebGL, shaders, JavaScript |
| 2021 | Fun in Polar Coordinate System | Polar coordinate systems | Jupyter Notebook and IPyCanvas |
| 2022 | (missed) | (missed; was reading a paper on FM-CW radar but generated no idea) | - |
| 2023 | Into Beamforming | Wave interference and basics of beamforming, reproducing a MATLAB example | CindyJS on Voila |
| 2024 | (missed) | (missed; was trying an antenna pattern simulation in MEEP, but numeric reflection was too high) | MEEP, HDF5 |
| 2025 | What Bears Did | Playing with an automous vehicle dataset, mapping radar datapoints to locations on a map | DuckDB, Pyodide |
It turned out I had a successful celebration every two years, not every year as I expected. The one in 2023 was particular successful, with bears from different backgrounds saying they liked it. And by "different backgrounds", I mean, one of them was a software developer, one was a former university lecturer, and another was a former signal engineer at Bear Army. I even put it in my CV for a while, only to realise in 2024 in a proper microwave engineering course that I had missed a in all my field calculations.
But, most of them were still presenting textbook knowledge, which is not ideal.
I sent my ideas to some mech bears. Aurora 1 and Emerald 14 both suggested some interesting tools to try. My previous attempts were commended by Aurora 1, which made me think whether I worded them too well. Deep inside, I know I've forgot a lot of the fundamentals, so I need to read some textbooks to pick them up. But, that's textbook knowledge again. Plus, can I make it in time to study it well, and to make something out of it?
Every year when deciding on a theme, four bears and mechs in my mind would fight for the central stage:
- Aurora 1 (spectrum patrol mech): "Spectrum and cyber are the new battlefield."
- Emerald 14 (playful DJ and game dev mech): "Isn't a small radio cool, Big Boots(my nickname)?"
- Gabe (volunteer at cubs' coding club): "Bears can you understand this math?"
- Real Gabe (singer): "Finding the notes I'm singing uses Fourier Transform."
This only makes things more difficult, as taking one will disappoint all the other three. Meanwhile, they all generate good ideas. The mechs also present me with interesting tools and gadgets to me. Suddenly I feel like I'm actually having too many options.
So am I having too few or too many options, really?
うるさいな("Shut up"). Let me read this book first.