Python where we can, C++ where we must
As a Embedded Software developer, if you are outside of the industry you will imagine that we always work with C and C++. However, for testing setups and automatization processes we always used Python. Why? Because it is easy to use, easy to implement and fast for prototyping.
Since Python is also the foundational language for Yocto, mastering it lets you tap into Yocto’s internal logic and modules. Here, I’m documenting the tools and Python patterns I find most useful—especially those "don't forget" notes for when I've been staring at C/C++ headers for too long and need to switch back to Python’s syntax.
AI Generated with Gemini Nano Banana Pro
Clean flat vector illustration, modern editorial tech style: crisp solid shapes, bold silhouettes, minimal gradients. A composition split exactly down the middle into two domains. Left half: a single friendly stylized flat snake working fluidly at a computer, conveying fast, easy prototyping and testing — open, airy, light. Right half: two stylized flat snakes handling tools, with a robotic arm carrying a microchip and a faint decorative field of ones-and-zeros texture behind them, conveying the rigid low-level "must" domain. Blue-dominant palette derived from deep navy (#0b1120), blue (#3b82f6, #60a5fa) and pale blue (#dbeafe), freely using shades and gradients, with cyan (#06b6d4) and a touch of teal as minority analogous accents (~10–25% of the frame); the snakes rendered in blue-family tones, not brand colors. A subtle vertical seam divides the two halves. Clear focal hierarchy, legible at thumbnail size, right half kept from getting cluttered. Absolutely no text, no letters, no readable numbers, no words, no labels. No brand logos. No human faces. Stylized flat animal/snake and simple flat icons are allowed. 16:9 aspect ratio.
Styling Guidelines for Python projects
For Python projects I tend to be on the side of PEP8. I follow their coding style and use yapf, which was developed by Google to control the styling guide in python projects. My file is fairly simple:
How to create a basic python module
Nowadays on times of python 3.14.X there are a tone of ways to create a python module. This is kind of the "old" way. You will be find python purist that say why this is no longer recommended, but I decide to document this version here because it is the current way of installing modules that works out-of-the-box with Yocto and with legacy tools. Besides the blog post Is setup.py deprecated? shows why this is still relevant and how it could be used to follow the latest trends.